type
stringclasses 1
value | name
stringlengths 14
183
| virtualsite_url
stringlengths 46
46
| speakers/authors
stringlengths 8
1.31k
| abstract
stringlengths 246
3.59k
|
|---|---|---|---|---|
Poster
|
A faster training algorithm for regression trees with linear leaves, and an analysis of its complexity
|
https://neurips.cc//virtual/2025/poster/115461
|
Kuat Gazizov, Miguel A. Carreira-Perpinan
|
We consider the Tree Alternating Optimization (TAO) algorithm to train regression trees with linear predictors in the leaves. Unlike the traditional, greedy recursive partitioning algorithms such as CART, TAO guarantees a monotonic decrease of the objective function and results in smaller trees of much better accuracy. We modify the TAO algorithm so that it produces exactly the same result but is much faster, particularly for high input dimensionality or deep trees. The idea is based on the fact that, at each iteration of TAO, each leaf receives only a subset of the training instances. Thus, the optimization of the leaf model can be done exactly but faster by using the Sherman-Morrison-Woodbury formula. This has the unexpected advantage that, once a tree exceeds a critical depth, then making it deeper makes it faster to train, even though the tree is larger and has more parameters. Indeed, this can make learning a nonlinear model (the tree) asymptotically faster than a regular linear regression model. We analyze the corresponding computational complexity and verify the speedups experimentally in various datasets. The argument can be applied to other types of trees, whenever the optimization of a node can be computed in superlinear time of the number of instances.
|
Poster
|
A Few Moments Please: Scalable Graphon Learning via Moment Matching
|
https://neurips.cc//virtual/2025/poster/117937
|
Reza Ramezanpour, Victor Tenorio, Antonio G. Marques, Ashutosh Sabharwal, Santiago Segarra
|
Graphons, as limit objects of dense graph sequences, play a central role in the statistical analysis of network data. However, existing graphon estimation methods often struggle with scalability to large networks and resolution-independent approximation, due to their reliance on estimating latent variables or costly metrics such as the Gromov-Wasserstein distance. In this work, we propose a novel, scalable graphon estimator that directly recovers the graphon via moment matching, leveraging implicit neural representations (INRs). Our approach avoids latent variable modeling by training an INR--mapping coordinates to graphon values--to match empirical subgraph counts (i.e., moments) from observed graphs.This direct estimation mechanism yields a polynomial-time solution and crucially sidesteps the combinatorial complexity of Gromov-Wasserstein optimization. Building on foundational results, we establish a theoretical guarantee: when the observed subgraph motifs sufficiently represent those of the true graphon (a condition met with sufficiently large or numerous graph samples), the estimated graphon achieves a provable upper bound in cut distance from the ground truth. Additionally, we introduce MomentMixup, a data augmentation technique that performs mixup in the moment space to enhance graphon-based learning. Our graphon estimation method achieves strong empirical performance--demonstrating high accuracy on small graphs and superior computational efficiency on large graphs--outperforming state-of-the-art scalable estimators in 75\% of benchmark settings and matching them in the remaining cases. Furthermore, MomentMixup demonstrated improved graph classification accuracy on the majority of our benchmarks.
|
Poster
|
Affine-Invariant Global Non-Asymptotic Convergence Analysis of BFGS under Self-Concordance
|
https://neurips.cc//virtual/2025/poster/117030
|
Qiujiang Jin, Aryan Mokhtari
|
In this paper, we establish global non-asymptotic convergence guarantees for the BFGS quasi-Newton method without requiring strong convexity or the Lipschitz continuity of the gradient or Hessian. Instead, we consider the setting where the objective function is strictly convex and strongly self-concordant. For an arbitrary initial point and any arbitrary positive-definite initial Hessian approximation, we prove global linear and superlinear convergence guarantees for BFGS when the step size is determined using a line search scheme satisfying the weak Wolfe conditions. Moreover, all our global guarantees are affine-invariant, with the convergence rates depending solely on the initial error and the strongly self-concordant constant. Our results extend the global non-asymptotic convergence theory of BFGS beyond traditional assumptions and, for the first time, establish affine-invariant convergence guarantees—aligning with the inherent affine invariance of the BFGS method.
|
Poster
|
AffordBot: 3D Fine-grained Embodied Reasoning via Multimodal Large Language Models
|
https://neurips.cc//virtual/2025/poster/118403
|
Xinyi Wang, Xun Yang, Yanlong Xu, Yuchen Wu, Zhen Li, Na Zhao
|
Effective human–agent collaboration in physical environments requires understanding not only what to act upon, but also where the actionable elements are and how to interact with them. Existing approaches often operate at the object level or disjointedly handle fine-grained affordance reasoning, lacking coherent, instruction-driven grounding and reasoning. In this work, we introduce a new task: Fine-grained 3D Embodied Reasoning, which requires an agent to predict, for each referenced affordance element in a 3D scene, a structured triplet comprising its spatial location, motion type, and motion axis, based on a task instruction. To solve this task, we propose AffordBot, a novel framework that integrates Multimodal Large Language Models (MLLMs) with a tailored chain-of-thought (CoT) reasoning paradigm. To bridge the gap between 3D input and 2D-compatible MLLMs, we render surround-view images of the scene and project 3D element candidates into these views, forming a rich visual representation aligned with the scene geometry. Our CoT pipeline begins with an active perception stage, prompting the MLLM to select the most informative viewpoint based on the instruction, before proceeding with step-by-step reasoning to localize affordance elements and infer plausible interaction motions. Evaluated on the SceneFun3D dataset, AffordBot achieves state-of-the-art performance, demonstrating strong generalization and physically grounded reasoning with only 3D point cloud input and MLLMs.
|
Poster
|
A Finite Sample Analysis of Distributional TD Learning with Linear Function Approximation
|
https://neurips.cc//virtual/2025/poster/120107
|
Yang Peng, Kaicheng Jin, Liangyu Zhang, Zhihua Zhang
|
In this paper, we study the finite-sample statistical rates of distributional temporal difference (TD) learning with linear function approximation.The aim of distributional TD learning is to estimate the return distribution of a discounted Markov decision process for a given policy $\pi$.Previous works on statistical analysis of distributional TD learning mainly focus on the tabular case.In contrast, we first consider the linear function approximation setting and derive sharp finite-sample rates.Our theoretical results demonstrate that the sample complexity of linear distributional TD learning matches that of classic linear TD learning.This implies that, with linear function approximation, learning the full distribution of the return from streaming data is no more difficult than learning its expectation (value function).To derive tight sample complexity bounds, we conduct a fine-grained analysis of the linear-categorical Bellman equation and employ the exponential stability arguments for products of random matrices.Our results provide new insights into the statistical efficiency of distributional reinforcement learning algorithms.
|
Poster
|
A Frustratingly Simple Yet Highly Effective Attack Baseline: Over 90% Success Rate Against the Strong Black-box Models of GPT-4.5/4o/o1
|
https://neurips.cc//virtual/2025/poster/119497
|
Zhaoyi Li, Xiaohan Zhao, Dong-Dong Wu, Jiacheng Cui, Zhiqiang Shen
|
Despite promising performance on open-source large vision-language models (LVLMs), transfer-based targeted attacks often fail against black-box commercial closed-source LVLMs. Analyzing failed adversarial perturbations reveals that the learned perturbations typically originate from a uniform distribution and lack clear semantic details, resulting in unintended responses. This critical absence of semantic information leads commercial LVLMs to either ignore the perturbation entirely or misinterpret its embedded semantics, thereby causing the attack to fail. To overcome these issues, we propose to refine semantic clarity by encoding explicit semantic details within local regions, thus ensuring interoperability and capturing finer-grained features, and by concentrating modifications on semantically rich areas rather than applying them uniformly. To achieve this, we propose *a simple yet highly effective baseline*: at each optimization step, the adversarial image is cropped randomly by a controlled aspect ratio and scale, resized, and then aligned with the target image in the embedding space. While the na\"ive source-target matching method has been utilized before in the literature, we are the first to provide a tight analysis, which establishes a close connection between perturbation optimization and semantics. Experimental results confirm our hypothesis. Our adversarial examples crafted with local-aggregated perturbations focused on crucial regions exhibit surprisingly good transferability to commercial LVLMs, including GPT-4.5, GPT-4o, Gemini-2.0-flash, Claude-3.5/3.7-sonnet, and even reasoning models like o1, Claude-3.7-thinking and Gemini-2.0-flash-thinking. Our approach achieves success rates exceeding 90\% on GPT-4.5, 4o, and o1, significantly outperforming all prior state-of-the-art attack methods. Our code and optimized adversarial examples are available in supplementary materials.
|
Poster
|
Afterburner: Reinforcement Learning Facilitates Self-Improving Code Efficiency Optimization
|
https://neurips.cc//virtual/2025/poster/119429
|
Mingzhe Du, Anh Tuan Luu, Yue Liu, Yuhao QING, Dong HUANG, Xinyi He, Qian Liu, Zejun MA, See-Kiong Ng
|
Large Language Models (LLMs) generate functionally correct solutions but often fall short in code efficiency, a critical bottleneck for real-world deployment. In this paper, we introduce a novel test-time iterative optimization framework to address this, employing a closed-loop system where LLMs iteratively refine code based on empirical performance feedback from an execution sandbox. We explore three training strategies: Supervised Fine-Tuning (SFT), Direct Preference Optimization (DPO), and Group Relative Policy Optimization~(GRPO). Experiments on our Venus dataset and the APPS benchmark show that SFT and DPO rapidly saturate in efficiency gains. In contrast, GRPO, using reinforcement learning (RL) with execution feedback, continuously optimizes code performance, significantly boosting both pass@1 (from 47% to 62%) and the likelihood of outperforming human submissions in efficiency (from 31% to 45%). Our work demonstrates effective test-time code efficiency improvement and critically reveals the power of RL in teaching LLMs to truly self-improve code efficiency.
|
Poster
|
AF-UMC: An Alignment-Free Fusion Framework for Unaligned Multi-View Clustering
|
https://neurips.cc//virtual/2025/poster/118997
|
Bohang Sun, Yuena Lin, Tao Yang, Zhen Zhu, Zhen Yang, Gengyu Lyu
|
The Unaligned Multi-view Clustering (UMC) aims to learn a discriminative cluster structure from unaligned multi-view data, where the features of samples are not completely aligned across multiple views. Most of current methods usually prioritize employing various alignment strategies to align different sample representations and then conduct cross-view fusion for subsequent clustering. However, ***due to the heterogeneity of representations across different views, these alignment strategies often fail to achieve ideal view-alignment results, inevitably leading to unreliable cross-view fusion.*** To address this issue, we propose an alignment-free consistency fusion framework named AF-UMC, which bypasses the traditional view-alignment operation and directly extracts consistent representation from each view to perform global cross-view consistency fusion. Specifically, we first construct a cross-view consistent basis space by a cross-view reconstruction loss and a designed Structural Clarity Regularization (SCR), where autoencoders extract consistent representation of each view through projecting view-specific data to the constructed basis space. Afterwards, these extracted representations are globally pulled together for further cross-view fusion according to a designed Instance Global Contrastive Enhancement (IGCE), which makes the fused consistent representation with higher global consistency. Compared with previous methods, AF-UMC directly extracts consistent representation from each view for global fusion rather than alignment for fusion, which significantly mitigates the degraded performance caused by undesired view-alignment results while greatly reduces algorithm complexity and enhances its efficiency. Extensive experiments on various datasets demonstrate that our proposed method exhibits superior performance against other state-of-the-art methods.
|
Poster
|
AGC-Drive: A Large-Scale Dataset for Real-World Aerial-Ground Collaboration in Driving Scenarios
|
https://neurips.cc//virtual/2025/poster/121689
|
侯 云浩, Bochao Zou, Min Zhang, 燃 陈, Shangdong Yang, Yanmei Zhang, Junbao Zhuo, Siheng Chen, Jiansheng Chen, Huimin Ma
|
By sharing information across multiple agents, collaborative perception helps autonomous vehicles mitigate occlusions and improve overall perception accuracy. While most previous work focus on vehicle-to-vehicle and vehicle-to-infrastructure collaboration, with limited attention to aerial perspectives provided by UAVs, which uniquely offer dynamic, top-down views to alleviate occlusions and monitor large-scale interactive environments. A major reason for this is the lack of high-quality datasets for aerial-ground collaborative scenarios. To bridge this gap, we present AGC-Drive, the first large-scale real-world dataset for Aerial-Ground Cooperative 3D perception. The data collection platform consists of two vehicles, each equipped with five cameras and one LiDAR sensor, and one UAV carrying a forward-facing camera and a LiDAR sensor, enabling comprehensive multi-view and multi-agent perception. Consisting of approximately 120k LiDAR frames and 440k images, the dataset covers 14 diverse real-world driving scenarios, including urban roundabouts, highway tunnels, and on/off ramps. Notably, 19.5% of the data comprises dynamic interaction events, including vehicle cut-ins, cut-outs, and frequent lane changes. AGC-Drive contains 400 scenes, each with approximately 100 frames and fully annotated 3D bounding boxes covering 13 object categories. We provide benchmarks for two 3D perception tasks: vehicle-to-vehicle collaborative perception and vehicle-to-UAV collaborative perception. Additionally, we release an open-source toolkit, including spatiotemporal alignment verification tools, multi-agent visualization systems, and collaborative annotation utilities. The dataset and code are available at https://github.com/PercepX/AGC-Drive.
|
Poster
|
A Generalist Intracortical Motor Decoder
|
https://neurips.cc//virtual/2025/poster/115457
|
Joel Ye, Fabio Rizzoglio, Xuan Ma, Adam Smoulder, Hongwei Mao, Gary Blumenthal, William Hockeimer, Nicolas Kunigk, Dalton Moore, Patrick Marino, Raeed Chowdhury, J. Patrick Mayo, Aaron Batista, Steven Chase, Michael Boninger, Charles Greenspon, Andrew B Schwartz, Nicholas Hatsopoulos, Lee Miller, Kristofer Bouchard, Jennifer Collinger, Leila Wehbe, Robert Gaunt
|
Mapping the relationship between neural activity and motor behavior is a central aim of sensorimotor neuroscience and neurotechnology. While most progress to this end has relied on restricting complexity, the advent of foundation models instead proposes integrating a breadth of data as an alternate avenue for broadly advancing downstream modeling. We quantify this premise for motor decoding from intracortical microelectrode data, pretraining an autoregressive Transformer on 2000 hours of neural population spiking activity paired with diverse motor covariates from over 30 monkeys and humans. The resulting model is broadly useful, benefiting decoding on 8 downstream decoding tasks and generalizing to a variety of neural distribution shifts. However, we also highlight that scaling autoregressive Transformers seems unlikely to resolve limitations stemming from sensor variability and output stereotypy in neural datasets.
|
Poster
|
A Generalized Binary Tree Mechanism for Private Approximation of All-Pair Shortest Distances
|
https://neurips.cc//virtual/2025/poster/115388
|
Zongrui Zou, Chenglin Fan, Michael Dinitz, Jingcheng Liu, Jalaj Upadhyay
|
We study the problem of approximating all-pair distances in a weighted undirected graph with differential privacy, introduced by Sealfon [Sea16]. Given a publicly known undirected graph, we treat the weights of edges as sensitive information, and two graphs are neighbors if their edge weights differ in one edge by at most one. We obtain efficient algorithms with significantly improved bounds on a broad class of graphs which we refer to as *recursively separable*. In particular, for any $n$-vertex $K_h$-minor-free graph, our algorithm achieve an additive error of $ \widetilde{O}(h(nW)^{1/3} ) $, where $ W $ represents the maximum edge weight; For grid graphs, the same algorithmic scheme achieve additive error of $ \widetilde{O}(n^{1/4}\sqrt{W}) $.Our approach can be seen as a generalization of the celebrated binary tree mechanism for range queries, as releasing range queries is equivalent to computing all-pair distances on a path graph. In essence, our approach is based on generalizing the binary tree mechanism to graphs that are *recursively separable*.
|
Poster
|
A Generalized Bisimulation Metric of State Similarity between Markov Decision Processes: From Theoretical Propositions to Applications
|
https://neurips.cc//virtual/2025/poster/117504
|
ZHENYU TAO, Wei Xu, Xiaohu You
|
The bisimulation metric (BSM) is a powerful tool for computing state similarities within a Markov decision process (MDP), revealing that states closer in BSM have more similar optimal value functions. While BSM has been successfully utilized in reinforcement learning (RL) for tasks like state representation learning and policy exploration, its application to multiple-MDP scenarios, such as policy transfer, remains challenging. Prior work has attempted to generalize BSM to pairs of MDPs, but a lack of rigorous analysis of its mathematical properties has limited further theoretical progress. In this work, we formally establish a generalized bisimulation metric (GBSM) between pairs of MDPs, which is rigorously proven with the three fundamental properties: GBSM symmetry, inter-MDP triangle inequality, and the distance bound on identical states. Leveraging these properties, we theoretically analyse policy transfer, state aggregation, and sampling-based estimation in MDPs, obtaining explicit bounds that are strictly tighter than those derived from the standard BSM. Additionally, GBSM provides a closed-form sample complexity for estimation, improving upon existing asymptotic results based on BSM. Numerical results validate our theoretical findings and demonstrate the effectiveness of GBSM in multi-MDP scenarios.
|
Poster
|
A Generalized Iterative Imputation Framework for Model Adaptation and Oracle Feature Utilization
|
https://neurips.cc//virtual/2025/poster/118575
|
Hao Wang, zhengnan li, Zhichao Chen, Xu Chen, Shuting He, Guangyi Liu, Haoxuan Li, Zhouchen Lin
|
Iterative imputation is a prevalent method for completing missing data, which involves iteratively imputing each feature by treating it as a target variable and predicting its missing values using the remaining features. However, existing iterative imputation methods exhibit two critical defects: (1) model misspecification, where a uniform parametric form of model is applied across different features, conflicting with heterogeneous data generation processes; (2) underutilization of fully observed features, where all features are treated as potentially missing, neglecting the valuable information in fully observed features.In this work, we propose Kernel Point Imputation (KPI), a bi-level optimization framework designed to address these issues. The inner-level optimization optimizes the model form for each feature in a reproducing kernel Hilbert space, mitigating model misspecification. The outer-level optimization leverages fully observed features as supervision signals to refine imputations. Extensive experiments on real-world datasets demonstrate that KPI consistently outperforms state-of-the-art imputation methods.
|
Poster
|
A Generalized Label Shift Perspective for Cross-Domain Gaze Estimation
|
https://neurips.cc//virtual/2025/poster/116912
|
Haoran Yang, Xiaohui Chen, Chuan-Xian Ren
|
Aiming to generalize the well-trained gaze estimation model to new target domains, Cross-domain Gaze Estimation (CDGE) is developed for real-world application scenarios. Existing CDGE methods typically extract the domain-invariant features to mitigate domain shift in feature space, which is proved insufficient by Generalized Label Shift (GLS) theory. In this paper, we introduce a novel GLS perspective to CDGE and modelize the cross-domain problem by label and conditional shift problem. A GLS correction framework is presented and a feasible realization is proposed, in which a importance reweighting strategy based on truncated Gaussian distribution is introduced to overcome the continuity challenges in label shift correction. To embed the reweighted source distribution to conditional invariant learning, we further derive a probability-aware estimation of conditional operator discrepancy. Extensive experiments on standard CDGE tasks with different backbone models validate the superior generalization capability across domain and applicability on various models of proposed method.
|
Poster
|
A General-Purpose Theorem for High-Probability Bounds of Stochastic Approximation with Polyak Averaging
|
https://neurips.cc//virtual/2025/poster/119445
|
Sajad Khodadadian, Martin Zubeldia
|
Polyak–Ruppert averaging is a widely used technique to achieve the optimal asymptotic variance of stochastic approximation (SA) algorithms, yet its high-probability performance guarantees remain underexplored in general settings. In this paper, we present a general framework for establishing non-asymptotic concentration bounds for the error of averaged SA iterates. Our approach assumes access to individual concentration bounds for the unaveraged iterates and yields a sharp bound on the averaged iterates. We also construct an example, showing the tightness of our result up to constant multiplicative factors. As direct applications, we derive tight concentration bounds for contractive SA algorithms and for algorithms such as temporal difference learning and $Q$-learning with averaging, obtaining new bounds in settings where traditional analysis is challenging.
|
Poster
|
AgentAuditor: Human-level Safety and Security Evaluation for LLM Agents
|
https://neurips.cc//virtual/2025/poster/120154
|
Hanjun Luo, Shenyu Dai, Chiming Ni, Xinfeng Li, Guibin Zhang, Kun Wang, Tongliang Liu, Hanan Salam
|
Despite the rapid advancement of LLM-based agents, the reliable evaluation of their safety and security remains a significant challenge. Existing rule-based or LLM-based evaluators often miss dangers in agents' step-by-step actions, overlook subtle meanings, fail to see how small issues compound, and get confused by unclear safety or security rules. To overcome this evaluation crisis, we introduce AgentAuditor, a universal, training-free, memory-augmented reasoning framework that empowers LLM evaluators to emulate human expert evaluators. AgentAuditor constructs an experiential memory by having an LLM adaptively extract structured semantic features (e.g., scenario, risk, behavior) and generate associated chain-of-thought reasoning traces for past interactions. A multi-stage, context-aware retrieval-augmented generation process then dynamically retrieves the most relevant reasoning experiences to guide the LLM evaluator's assessment of new cases. Moreover, we developed ASSEBench, the first benchmark designed to check how well LLM-based evaluators can spot both safety risks and security threats. ASSEBench comprises 2293 meticulously annotated interaction records, covering 15 risk types across 29 application scenarios. A key feature of ASSEBench is its nuanced approach to ambiguous risk situations, employing "Strict" and "Lenient" judgment standards. Experiments demonstrate that AgentAuditor not only consistently improves the evaluation performance of LLMs across all benchmarks but also sets a new state-of-the-art in LLM-as-a-judge for agent safety and security, achieving human-level accuracy. Our work is openly accessible.
|
Poster
|
AgentBreeder: Mitigating the AI Safety Impact of Multi-Agent Scaffolds via Self-Improvement
|
https://neurips.cc//virtual/2025/poster/116189
|
J Rosser, Jakob Foerster
|
Scaffolding Large Language Models (LLMs) into multi-agent systems often improves performance on complex tasks, but the safety impact of such scaffolds has not been thoroughly explored. We introduce AgentBreeder, a framework for multi-objective self-improving evolutionary search over scaffolds. We evaluate discovered scaffolds on widely recognized reasoning, mathematics, and safety benchmarks and compare them with popular baselines. In 'blue' mode, we see a 79.4% average uplift in safety benchmark performance while maintaining or improving capability scores. In 'red' mode, we find adversarially weak scaffolds emerging concurrently with capability optimization. Our work demonstrates the risks of multi-agent scaffolding and provides a framework for mitigating them. Code is available at https://anonymous.4open.science/r/AgentBreeder-86AF .
|
Poster
|
AgentDAM: Privacy Leakage Evaluation for Autonomous Web Agents
|
https://neurips.cc//virtual/2025/poster/121443
|
Arman Zharmagambetov, Chuan Guo, Ivan Evtimov, Maya Pavlova, Ruslan Salakhutdinov, Kamalika Chaudhuri
|
Autonomous AI agents that can follow instructions and perform complex multi-step tasks have tremendous potential to boost human productivity. However, to perform many of these tasks, the agents need access to personal information from their users, raising the question of whether they are capable of using it appropriately. In this work, we introduce a new benchmark **AgentDAM** that measures if AI web-navigation agents follow the privacy principle of *"data minimization"*. For the purposes of our benchmark, data minimization means that the agent uses a piece of potentially sensitive information only if it is "necessary" to complete a particular task. Our benchmark simulates realistic web interaction scenarios end-to-end and is adaptable to all existing web navigation agents. We use AgentDAM to evaluate how well AI agents built on top of GPT-4, Llama-3 and Claude can limit processing of potentially private information, and show that they are prone to inadvertent use of unnecessary sensitive information. We also propose a prompting-based defense that reduces information leakage, and demonstrate that our end-to-end benchmarking provides a more realistic measure than probing LLMs about privacy. Our results highlight that further research is needed to develop AI agents that can prioritize data minimization at inference time.
|
Poster
|
AGENTIF: Benchmarking Large Language Models Instruction Following Ability in Agentic Scenarios
|
https://neurips.cc//virtual/2025/poster/121761
|
Yunjia Qi, Hao Peng, Xiaozhi Wang, Amy Xin, Youfeng Liu, Bin Xu, Lei Hou, Juanzi Li
|
Large Language Models (LLMs) have demonstrated advanced capabilities in real-world agentic applications. Growing research efforts aim to develop LLM-based agents to address practical demands, introducing a new challenge: agentic scenarios often involve lengthy instructions with complex constraints, such as extended system prompts and detailed tool specifications. While adherence to such instructions is crucial for agentic applications, whether LLMs can reliably follow them remains underexplored. In this paper, we introduce AgentIF, the first benchmark for systematically evaluating LLM instruction following ability in agentic scenarios. AgentIF features three key characteristics: (1) Realistic, constructed from $50$ real-world agentic applications. (2) Long, averaging $1,723$ words with a maximum of $15,630$ words. (3) Complex, averaging $11.9$ constraints per instruction, covering diverse constraint types, such as tool specifications and condition constraints.To construct AgentIF, we collect $707$ human-annotated instructions across $50$ agentic tasks from industrial application agents and open-source agentic systems. For each instruction, we annotate the associated constraints and corresponding evaluation metrics, including code-based evaluation, LLM-based evaluation, and hybrid code-LLM evaluation.% resulting in a final set of $707$ instructions.We use AgentIF to systematically evaluate existing advanced LLMs. We observe that current models generally perform poorly, especially in handling complex constraint structures and tool specifications. We further conduct error analysis and analytical experiments on instruction length and meta constraints, providing some findings about the failure modes of existing LLMs. We have released the code and data to facilitate future research.
|
Poster
|
AgentNet: Decentralized Evolutionary Coordination for LLM-based Multi-Agent Systems
|
https://neurips.cc//virtual/2025/poster/115584
|
Yingxuan Yang, Huacan Chai, Shuai Shao, Yuanyi Song, Siyuan Qi, Renting Rui, Weinan Zhang
|
The rapid advancement of Large Language Models (LLMs) has catalyzed the development of multi-agent systems, where multiple LLM-based agents collaborate to solve complex tasks. However, existing systems predominantly rely on centralized coordination, which introduces scalability bottlenecks, limits adaptability, and creates single points of failure. Additionally, concerns over privacy and proprietary knowledge sharing hinder cross-organizational collaboration, leading to siloed expertise. To address these challenges, we propose AgentNet, a decentralized, Retrieval-Augmented Generation (RAG)-based framework that enables LLM-based agents to autonomously evolve their capabilities and collaborate efficiently in a Directed Acyclic Graph (DAG)-structured network. Unlike traditional multi-agent systems that depend on static role assignments or centralized control, AgentNet allows agents to specialize dynamically, adjust their connectivity, and route tasks without relying on predefined workflows.AgentNet’s core design is built upon several key innovations: (1) Fully Decentralized Paradigm: Removing the central orchestrator, allowing agents to coordinate and specialize autonomously, fostering fault tolerance and emergent collective intelligence. (2) Dynamically Evolving Graph Topology: Real-time adaptation of agent connections based on task demands, ensuring scalability and resilience.(3) Adaptive Learning for Expertise Refinement: A retrieval-based memory system that enables agents to continuously update and refine their specialized skills.By eliminating centralized control, AgentNet enhances fault tolerance, promotes scalable specialization, and enables privacy-preserving collaboration across organizations. Through decentralized coordination and minimal data exchange, agents can leverage diverse knowledge sources while safeguarding sensitive information. Experimental results demonstrate that AgentNet outperforms traditional centralized multi-agent systems, significantly improving efficiency, adaptability, and scalability in dynamic environments, making it a promising foundation for next-generation autonomous, privacy-respecting multi-agent ecosystems.
|
Poster
|
AgentNet: Open Foundations for Computer-Use Agents
|
https://neurips.cc//virtual/2025/poster/119771
|
Xinyuan Wang, Bowen Wang, Dunjie Lu, Junlin Yang, Tianbao Xie, Junli Wang, Jiaqi Deng, Xiaole Guo, Zhennan Shen, Zhuokai Li, Ryan Li, Xiaochuan Li, Junda Chen, Boyuan Zheng, LI PEIHANG, Fangyu Lei, Chen Wu, Ruisheng Cao, Yeqiao Fu, Dongchan Shin, Martin Shin, Hu Jiarui, Yuyan Wang, Jixuan Chen, Yuxiao Ye, Yiheng Xu, Danyang Zhang, Yipu Wang, Heng Wang, Diyi Yang, Victor Zhong, Y.Charles, Zhilin Yang, Tao Yu
|
Vision-language models have demonstrated impressive capabilities as computer-use agents (CUAs) capable of automating diverse computer tasks. As their commercial potential grows, critical details of the most capable CUA systems remain closed and proprietary. As these agents will increasingly mediate digital interactions and execute consequential decisions on our behalf, the research community needs access to truly open CUA frameworks to study their capabilities, limitations, and risks. To bridge this gap, we propose AgentNet, a comprehensive open-source framework for scaling CUA data and foundation models. Our framework consists of: (1) an annotation infrastructure that seamlessly captures human computer-use demonstrations; (2) AgentNet dataset, a dataset of 27K computer-use data samples spanning various operating systems, applications, and websites; (3) a pipeline that discretizes continuous actions into state-action pairs and synthesizes reflective long chain-of-thought (CoT) reasoning; (4) a training recipe for scalable CUA modeling; and (5) AgentNetBench, a multi-dimensional offline benchmark for faster CUA evaluation. Our AgentNet-7B, fine-tuned on AgentNet dataset, demonstrates strong performance on several CUA benchmarks, achieving a success rate of 20.1% on OSWorld and 21.1% on WindowsAgentArena. Our training recipe, particularly its advanced reasoning mechanisms and strategic data mixture, enables robust performance scaling with increased data size. Further in-depth analysis of our models also demonstrate strong cross-domain generalization and performance scaling with test-time compute. We will release the annotation tool, datasets, code, and models to build open foundations for further CUA research.
|
Poster
|
AgentRecBench: Benchmarking LLM Agent-based Personalized Recommender Systems
|
https://neurips.cc//virtual/2025/poster/121525
|
Yu Shang, Peijie Liu, Yuwei Yan, Zijing Wu, Leheng Sheng, Yuanqing Yu, Chumeng Jiang, An Zhang, Fengli Xu, Yu Wang, Min Zhang, Yong Li
|
The emergence of agentic recommender systems powered by Large Language Models (LLMs) represents a paradigm shift in personalized recommendations, leveraging LLMs' advanced reasoning and role-playing capabilities to enable autonomous, adaptive decision-making. Unlike traditional recommendation approaches, agentic recommender systems can dynamically gather and interpret user-item interactions from complex environments, generating robust recommendation strategies that generalize across diverse scenarios. However, the field currently lacks standardized evaluation protocols to systematically assess these methods. To address this critical gap, we propose: (1) an interactive textual recommendation simulator incorporating rich user and item metadata and three typical evaluation scenarios (classic, evolving-interest, and cold-start recommendation tasks); (2) a unified modular framework for developing and studying agentic recommender systems; and (3) the first comprehensive benchmark comparing 10 classical and agentic recommendation methods. Our findings demonstrate the superiority of agentic systems and establish actionable design guidelines for their core components. The benchmark environment has been rigorously validated through an open challenge and remains publicly available with a continuously maintained leaderboard, fostering ongoing community engagement and reproducible research.The benchmark is available at: https://huggingface.co/datasets/SGJQovo/AgentRecBench.
|
Poster
|
Agent RL Scaling Law: Spontaneous Code Execution for Mathematical Problem Solving
|
https://neurips.cc//virtual/2025/poster/116372
|
Xinji Mai, Haotian Xu, Xing W, Weinong Wang, Yingying Zhang, Wenqiang Zhang
|
Large Language Models (LLMs) often struggle with mathematical reasoning tasks requiring precise, verifiable computation. While Reinforcement Learning (RL) from outcome-based rewards enhances text-based reasoning, understanding how agents autonomously learn to leverage external tools like code execution remains crucial. We investigate RL from outcome-based rewards for Tool-Integrated Reasoning, ZeroTIR, training base LLMs to spontaneously generate and execute Python code for mathematical problems without supervised tool-use examples. Our central contribution is we demonstrate that as RL training progresses, key metrics scale predictably. Specifically, we observe strong positive correlations where increased training steps lead to increases in the spontaneous code execution frequency, the average response length, and, critically, the final task accuracy. This suggests a quantifiable relationship between computational effort invested in training and the emergence of effective, tool-augmented reasoning strategies. We implement a robust framework featuring a decoupled code execution environment and validate our findings across standard RL algorithms and frameworks. Experiments show ZeroTIR significantly surpasses non-tool ZeroRL baselines on challenging math benchmarks. Our findings provide a foundational understanding of how autonomous tool use is acquired and scales within Agent RL, offering a reproducible benchmark for future studies. Code is released at \href{https://github.com/Anonymize-Author/AgentRL}{https://github.com/Anonymize-Author/AgentRL}.
|
Poster
|
Agents Robust to Distribution Shifts Learn Causal World Models Even Under Mediation
|
https://neurips.cc//virtual/2025/poster/118687
|
Matteo Ceriscioli, Karthika Mohan
|
Recent work [Richens and Everitt, 2024] has shown that agents robust to distribution shifts learn a causal model of their environment. However, these results rely on the assumption of no mediation, i.e., that an agent's actions do not affect their environment, which can be restrictive in many real-world settings. For example, a robot in an industrial plant might interact with tools, move through space, and transform products to complete its task. In this work, we extend the theoretical foundations of robust agency by proving that agents capable of adapting to distribution shifts must learn the underlying causal relationships even in the presence of mediation. We introduce an algorithm for eliciting Causal Influence Diagrams from robust agents using optimal policy oracles, with the flexibility to incorporate prior causal knowledge and demonstrate its effectiveness in mediated single-agent scenarios and multi-agent environments. We identify conditions under which the presence of a single robust agent is sufficient to recover the full causal model and derive optimal policies for other agents in the same environment. Finally, we demonstrate how to apply these results to sequential decision-making tasks modeled as Partially Observable Markov Decision Processes (POMDPs).
|
Poster
|
AgentTTS: Large Language Model Agent for Test-time Compute-optimal Scaling Strategy in Complex Tasks
|
https://neurips.cc//virtual/2025/poster/119334
|
Fali Wang, Hui Liu, Zhenwei Dai, Jingying Zeng, Zhiwei Zhang, Zongyu Wu, Chen Luo, Zhen Li, Xianfeng Tang, Qi He, Suhang Wang
|
Test-time scaling (TTS) enhances the performance of large language models (LLMs) by allocating additional compute resources during inference. However, existing research primarily investigates TTS in single-stage tasks; while many real-world problems are multi-stage complex tasks, composed of a sequence of heterogeneous subtasks with each subtask requires LLM of specific capability. Therefore, we study a novel problem: the test-time compute-optimal scaling in multi-stage complex tasks, aiming to select suitable models and allocate budgets per subtask to maximize overall performance. TTS in multi-stage tasks introduces two fundamental challenges: (i) The combinatorial search space of model and budget allocations, combined with the high cost of inference, makes brute-force search impractical. (ii) The optimal model and budget allocations across subtasks are interdependent, increasing the complexity of the compute-optimal search. To address this gap, we conduct extensive pilot experiments on four tasks across six datasets, deriving three empirical insights characterizing the behavior of LLMs in multi-stage complex tasks. Informed by these insights, we propose AgentTTS, an LLM-agent-based framework that autonomously searches for compute-optimal allocations through iterative feedback-driven interactions with the execution environment. Experimental results demonstrate that AgentTTS significantly outperforms traditional and other LLM-based baselines in search efficiency, and shows improved robustness to varying training set sizes and enhanced interpretability.
|
Poster
|
A geometric framework for momentum-based optimizers for low-rank training
|
https://neurips.cc//virtual/2025/poster/117118
|
Steffen Schotthöfer, Timon Klein, Jonas Kusch
|
Low-rank pre-training and fine-tuning have recently emerged as promising techniques for reducing the computational and storage costs of large neural networks. Training low-rank parameterizations typically relies on conventional optimizers such as heavy ball momentum methods or Adam. In this work, we identify and analyze potential difficulties that these training methods encounter when used to train low-rank parameterizations of weights. In particular, we show that classical momentum methods can struggle to converge to a local optimum due to the geometry of the underlying optimization landscape. To address this, we introduce novel training strategies derived from dynamical low-rank approximation, which explicitly account for the underlying geometric structure. Our approach leverages and combines tools from dynamical low-rank approximation and momentum-based optimization to design optimizers that respect the intrinsic geometry of the parameter space. We validate our methods through numerical experiments, demonstrating faster convergence, and stronger validation metrics at given parameter budgets.
|
Poster
|
A Geometry-Aware Metric for Mode Collapse in Multivariate Time Series Generative Models
|
https://neurips.cc//virtual/2025/poster/117444
|
Yassine ABBAHADDOU, Amine Aboussalah
|
Generative models such as Generative Adversarial Networks (GANs), Variational Autoencoders (VAEs), and diffusion models often fail to capture the full diversity of their training data, leading to mode collapse. While this issue is well-explored in image generation, it remains underinvestigated for time series data. We introduce a new definition of mode collapse specific to time series and propose a new geometry-aware metric, DMD-GEN, to quantify its severity. Our metric leverages Dynamic Mode Decomposition (DMD), a data-driven technique for identifying coherent spatiotemporal patterns through spectral analysis, and employs optimal transport between DMD eigenvectors to assess discrepancies in the underlying dynamics of the original and generated data. The geometry-aware nature of our method comes from modeling DMD modes as points on a Grassmann manifold and comparing them using Wasserstein distances computed via principal angles. When using mini-batch evaluation, DMD-GEN compares the subspaces spanned by the dominant modes of original and generated time series through optimal transport of their corresponding subspaces, enabling a principled geometric comparison. DMD-GEN is efficient in practice: it is used only during evaluation, supports mini-batch approximations, and is highly parallelizable. It not only quantifies the preservation of essential dynamic characteristics but also provides interpretability by highlighting which modes are poorly captured in the generated data. We validate DMD-GEN on both synthetic and real-world datasets using TimeGAN, TimeVAE, and DiffusionTS. The results show that DMD-GEN aligns well with traditional metrics while, for the first time, offering a principled definition of mode collapse for time series.
|
Poster
|
Aggregation Hides OOD Generalization Failures from Spurious Correlations
|
https://neurips.cc//virtual/2025/poster/115351
|
Olawale Salaudeen, Haoran Zhang, Kumail Alhamoud, Sara Beery, Marzyeh Ghassemi
|
Benchmarks for out‑of‑distribution (OOD) generalization frequently show a strong positive correlation between in‑distribution (ID) and OOD accuracy, termed "accuracy‑on‑the‑line." This pattern is often taken to imply that spurious correlations---correlations that improve ID but reduce OOD performance---are rare in practice. We find that this positive correlation is an artifact of aggregating heterogeneous OOD examples. Using a simple gradient‑based method, we partition each benchmark’s OOD split into semantically coherent subsets where accuracy on the line does not hold. Across six widely used distribution shift benchmarks, the method uncovers subsets, sometimes up to 77% of the usual OOD split, where higher ID accuracy predicts lower OOD accuracy. Our findings indicate that aggregate metrics can obscure important failure modes of OOD robustness. We release code and the identified subsets to facilitate further evaluation.
|
Poster
|
AGI-Elo: How Far Are We From Mastering A Task?
|
https://neurips.cc//virtual/2025/poster/121514
|
Shuo Sun, Yimin Zhao, Christina Lee, JIAWEI SUN, Chengran Yuan, Zefan Huang, Dongen Li, Justin Yeoh, Alok Prakash, Thomas Malone, Marcelo Ang Jr
|
As the field progresses toward Artificial General Intelligence (AGI), there is a pressing need for more comprehensive and insightful evaluation frameworks that go beyond aggregate performance metrics. This paper introduces a unified rating system that jointly models the difficulty of individual test cases and the competency of AI models (or humans) across vision, language, and action domains. Unlike existing metrics that focus solely on models, our approach allows for fine-grained, difficulty-aware evaluations through competitive interactions between models and tasks, capturing both the long-tail distribution of real-world challenges and the competency gap between current models and full task mastery. We validate the generalizability and robustness of our system through extensive experiments on multiple established datasets and models across distinct AGI domains. The resulting rating distributions offer novel perspectives and interpretable insights into task difficulty, model progression, and the outstanding challenges that remain on the path to achieving full AGI task mastery. We have made our code and results publicly available at https://ss47816.github.io/AGI-Elo/.
|
Poster
|
AgMMU: A Comprehensive Agricultural Multimodal Understanding Benchmark
|
https://neurips.cc//virtual/2025/poster/121696
|
Aruna Gauba, Irene Pi, Yunze Man, Ziqi Pang, Vikram Adve, Yu-Xiong Wang
|
We present **AgMMU**, a challenging real‑world benchmark for evaluating and advancing vision-language models (VLMs) in the knowledge‑intensive domain of agriculture. Unlike prior datasets that rely on crowdsourced prompts, AgMMU is distilled from 116,231 authentic dialogues between everyday growers and USDA-authorized Cooperative Extension experts. Through a three‑stage pipeline: automated knowledge extraction, QA generation, and human verification, we construct (i) AgMMU, an evaluation set of 746 multiple‑choice questions (MCQs) and 746 open‑ended questions (OEQs), and (ii) AgBase, a development corpus of 57,079 multimodal facts covering five high-stakes agricultural topics: insect identification, species identification, disease categorization, symptom description, and management instruction. AgMMU has three key advantages:- **Authentic \& Expert‑Verified**: All facts, images, and answers originate from real farmer and gardener inquiries answered by credentialed specialists, ensuring high‑fidelity agricultural knowledge.- **Complete Development Suite**: AgMMU uniquely couples a dual‑format evaluation benchmark (MCQ and OEQ) with AgBase, a large‑scale training set, enabling both rigorous assessment and targeted improvement of VLMs.- **Knowledge‑intensive Challenge**: Our tasks demand the synergy of nuanced visual perception and domain expertise, exposing fundamental limitations of current general‑purpose models and charting a path toward robust, application‑ready agricultural AI.Benchmarking 12 leading VLMs reveals pronounced gaps in fine‑grained perception and factual grounding. Open‑sourced models trail after proprietary ones by a wide margin. Simple fine‑tuning on AgBase boosts open-sourced model performance on challenging OEQs for up to 11.6\% on average, narrowing this gap and also motivating future research to propose better strategies in knowledge extraction and distillation from AgBase. We hope AgMMU stimulates research on domain‑specific knowledge integration and trustworthy decision support in agriculture AI development.
|
Poster
|
Agnostic Active Learning is Always Better Than Passive Learning
|
https://neurips.cc//virtual/2025/poster/117511
|
Steve Hanneke
|
We sharply characterize the optimal first-order query complexity of agnostic active learning for all concept classes, and propose a new general active learning algorithm which achieves it. Remarkably, the optimal query complexity admits a leading term which is always strictly smaller than the sample complexity of passive supervised learning (by a factor proportional to the best-in-class error rate). This was not previously known to be possible in the agnostic setting. For comparison, in all previous general analyses, the leading term exhibits an additional factor, such as the disagreement coefficient or related complexity measure, and therefore only provides improvements over passive learning in restricted cases. The present work completely removes such factors from the leading term, implying that $\textit{every}$ concept class benefits from active learning in the non-realizable case. The results established in this work resolve an important long-standing open question central to the past two decades of research on the theory of agnostic active learning.
|
Poster
|
Agnostic Continuous-Time Online Learning
|
https://neurips.cc//virtual/2025/poster/119547
|
Pramith Devulapalli, Changlong Wu, Ananth Grama, Wojciech Szpankowski
|
We study agnostic online learning from continuous-time data streams, a setting that naturally arises in applications such as environmental monitoring, personalized recommendation, and high-frequency trading. Unlike classical discrete-time models, learners in this setting must interact with a continually evolving data stream while making queries and updating models only at sparse, strategically selected times. We develop a general theoretical framework for learning from both *oblivious* and *adaptive* data streams, which may be noisy and non-stationary. For oblivious streams, we present a black-box reduction to classical online learning that yields a regret bound of $T \cdot R(S)/S$ for any class with discrete-time regret $R(S)$, where $T$ is the time horizon and $S$ is the *query budget*. For adaptive streams, which can evolve in response to learner actions, we design a dynamic query strategy in conjunction with a novel importance weighting scheme that enables unbiased loss estimation. In particular, for hypothesis class $\mathcal{H}$ with a finite Littlestone dimension, we establish a tight regret bound of $\tilde{\Theta}(T \cdot \sqrt{\mathsf{Ldim}(\mathcal{H})/S})$ that holds in both settings. Our results provide the first *quantitative* characterization of agnostic learning in continuous-time online environments with limited interaction.
|
Poster
|
Agnostic Learning under Targeted Poisoning: Optimal Rates and the Role of Randomness
|
https://neurips.cc//virtual/2025/poster/118672
|
Tom Waknine, Shay Moran, Bogdan Chornomaz, Yonatan Koren
|
We study the problem of learning in the presence of an adversary that can corrupt an $\eta$ fraction of the training examples with the goal of causing failure on a specific test point. In the realizable setting, prior work established that the optimal error under such instance-targeted poisoning attacks scales as $\Theta(d\eta)$, where $d$ is the VC dimension of the hypothesis class [Hanneke, Karbasi, Mahmoody, Mehalel, and Moran (NeurIPS 2022)]. In this work, we resolve the corresponding question in the agnostic setting. We show that the optimal excess error is $\widetilde\Theta(\sqrt{d\eta})$, answering one of the main open problems left by Hanneke et al. To achieve this rate, it is necessary to use randomized learners: Hanneke et al.\ showed that deterministic learners can be forced to suffer error close to $1$ even under small amounts of poisoning. Perhaps surprisingly, our upper bound remains valid even when the learner’s random bits are fully visible to the adversary. In the other direction, our lower bound is stronger than standard PAC-style bounds: instead of tailoring a hard distribution separately for each sample size, we exhibit a single fixed distribution under which the adversary can enforce an excess error of $\Omega(\sqrt{d\eta})$ infinitely often.
|
Poster
|
A Gradient Guidance Perspective on Stepwise Preference Optimization for Diffusion Models
|
https://neurips.cc//virtual/2025/poster/117028
|
Joshua Tian Jin Tee, Hee Suk Yoon, Abu Hanif Muhammad Syarubany, Eunseop Yoon, Chang Yoo
|
Direct Preference Optimization (DPO) is a key framework for aligning text-to-image models with human preferences, extended by Stepwise Preference Optimization (SPO) to leverage intermediate steps for preference learning, generating more aesthetically pleasing images with significantly less computational cost. While effective, SPO's underlying mechanisms remain underexplored. In light of this, We critically re-examine SPO by formalizing its mechanism as gradient guidance. This novel lens reveals SPO’s biased temporal weighting—underweighting later generative steps—and, uniquely compared to likelihood-centric views, highlights the presence of significant noise in these gradient estimates. Leveraging these insights, our GradSPO algorithm introduces a simplified loss and a targeted, variance-informed noise reduction strategy, enhancing training stability. Evaluations on SD 1.5 and SDXL show GradSPO substantially outperforms leading baselines in human preference, yielding images with markedly improved aesthetics and semantic faithfulness, leading to more robust alignment.
|
Poster
|
A Gradient Guided Diffusion Framework for Chance Constrained Programming
|
https://neurips.cc//virtual/2025/poster/119015
|
Boyang Zhang, Zhiguo Wang, Ya-Feng Liu
|
Chance constrained programming (CCP) is a powerful framework for addressing optimization problems under uncertainty. In this paper, we introduce a novel \textbf{G}radient-\textbf{G}uided \textbf{D}iffusion-based \textbf{Opt}imization framework, termed GGDOpt, which tackles CCP through three key innovations. First, GGDOpt accommodates a broad class of CCP problems without requiring the knowledge of the exact distribution of uncertainty—relying solely on a set of samples. Second, to address the nonconvexity of the chance constraints, it reformulates the CCP as a sampling problem over the product of two distributions: an unknown data distribution supported on a nonconvex set and a Boltzmann distribution defined by the objective function, which fully leverages both first- and second-order gradient information. Third, GGDOpt has theoretical convergence guarantees and provides practical error bounds under mild assumptions. By progressively injecting noise during the forward diffusion process to convexify the nonconvex feasible region, GGDOpt enables guided reverse sampling to generate asymptotically optimal solutions. Experimental results on synthetic datasets and a waveform design task in wireless communications demonstrate that GGDOpt outperforms existing methods in both solution quality and stability with nearly 80\% overhead reduction.
|
Poster
|
AHa-Bench: Benchmarking Audio Hallucinations in Large Audio-Language Models
|
https://neurips.cc//virtual/2025/poster/121405
|
Xize Cheng, Dongjie Fu, Chenyuhao Wen, Shannon Yu, Zehan Wang, Shengpeng Ji, Siddhant Arora, Tao Jin, Shinji Watanabe, Zhou Zhao
|
Hallucinations present a significant challenge in the development and evaluation of large language models (LLMs), directly affecting their reliability and accuracy. While notable advancements have been made in research on textual and visual hallucinations, there is still a lack of a comprehensive benchmark for evaluating auditory hallucinations in large audio language models (LALMs). To fill this gap, we introduce **AHa-Bench**, a systematic and comprehensive benchmark for audio hallucinations. Audio data, in particular, uniquely combines the multi-attribute complexity of visual data with the semantic richness of textual data, leading to auditory hallucinations that share characteristics with both visual and textual hallucinations. Based on the source of these hallucinations, AHa-Bench categorizes them into semantic hallucinations, acoustic hallucinations, and semantic-acoustic confusion hallucinations. In addition, we systematically evaluate seven open-source local perception language models (LALMs), demonstrating the challenges these models face in audio understanding, especially when it comes to jointly understanding semantic and acoustic information. Through the development of a comprehensive evaluation framework, AHa-Bench aims to enhance the robustness and stability of LALMs, fostering more reliable and nuanced audio understanding in LALMs. The benchmark dataset is available at \url{https://huggingface.co/datasets/ahabench/AHa-Bench}.
|
Poster
|
Aha! - Predicting What Matters Next: Online Highlight Detection Without Looking Ahead
|
https://neurips.cc//virtual/2025/poster/119707
|
Aiden Chang, Celso de Melo, Stephanie Lukin
|
Real-time understanding of continuous video streams is essential for intelligent agents operating in high-stakes environments, including autonomous vehicles, surveillance drones, and disaster response robots. Yet, most existing video understanding and highlight detection methods assume access to the entire video during inference, making them unsuitable for online or streaming scenarios. In particular, current models optimize for offline summarization, failing to support step-by-step reasoning needed for real-time decision-making. We introduce Aha!, an autoregressive highlight detection framework that predicts the relevance of each video frame against a task described in natural language. Without accessing future video frames, Aha! utilizes a multimodal language-vision model and lightweight, decoupled heads trained on a large, curated dataset of human-centric video labels. To enable scalability, we adopt a fixed-size SinkCache mechanism that achieves constant memory usage across infinite-length streams without degrading performance on standard benchmarks. This encourages the hidden representation to capture high-level task objectives, enabling effective frame-level rankings for informativeness, relevance, and uncertainty with respect to the natural language task. Aha! achieves state-of-the-art performance on highlight detection benchmarks, surpassing prior full-context and video-language models by +5.5\% on TVSum and +8.3\% on Mr. HiSum in mAP. We explore Aha!’s potential for real-world robotics applications given a task-oriented natural language input and a continuous, robot-centric video. Both experiments demonstrate Aha!'s potential effectiveness as a real-time reasoning module for downstream planning and long-horizon understanding.
|
Poster
|
A Hierarchy of Graphical Models for Counterfactual Inferences
|
https://neurips.cc//virtual/2025/poster/115508
|
Hongshuo Yang, Elias Bareinboim
|
Graphical models have been widely used as parsimonious encoders of assumptions of the underlying structural causal models and provide a basis from which causal inferences can be performed. Models that encode stronger constraints tend to have higher expressive power at the expense of lower empirical falsifiability. In this paper, we define two new collections of distributions which include counterfactual quantities that become experimentally accessible under the counterfactual randomization action. Correspondingly, we provide two new classes of graphical models for encoding empirically testable constraints in these distributions. We further present a sound and complete calculus, based on counterfactual calculus, which licenses inference in these two new models with rules that are also fall within the empirically falsifiable boundary. In addition, we formulate a hierarchy over several graphical models based on the constraints they encode and study the fundamental trade-off between the expressive power and empirical falsifiability of different models across the hierarchy.
|
Poster
|
A High-Dimensional Statistical Method for Optimizing Transfer Quantities in Multi-Source Transfer Learning
|
https://neurips.cc//virtual/2025/poster/116795
|
Qingyue Zhang, Haohao Fu, Guanbo Huang, Yaoyuan Liang, Chang Chu, Tianren Peng, Yanru Wu, Qi Li, Yang Li, Shao-Lun Huang
|
Multi-source transfer learning provides an effective solution to data scarcity in real-world supervised learning scenarios by leveraging multiple source tasks. In this field, existing works typically use all available samples from sources in training, which constrains their training efficiency and may lead to suboptimal results. To address this, we propose a theoretical framework that answers the question: what is the optimal quantity of source samples needed from each source task to jointly train the target model? Specifically, we introduce a generalization error measure based on K-L divergence, and minimize it based on high-dimensional statistical analysis to determine the optimal transfer quantity for each source task. Additionally, we develop an architecture-agnostic and data-efficient algorithm OTQMS to implement our theoretical results for target model training in multi-source transfer learning. Experimental studies on diverse architectures and two real-world benchmark datasets show that our proposed algorithm significantly outperforms state-of-the-art approaches in both accuracy and data efficiency. The code is available at https://anonymous.4open.science/r/Materials.
|
Poster
|
A Highly Efficient and Chemical Motif-Preserving Molecule Generation Platform
|
https://neurips.cc//virtual/2025/poster/119825
|
Peizhi Niu, Yu-Hsiang Wang, Vishal Rana, Chetan Rupakheti, Abhishek Pandey, Olgica Milenkovic
|
We introduce a new graph diffusion model for small drug molecule generation which simultaneously offers a $10$-fold reduction in the number of diffusion steps when compared to other existing methods, preservation of small molecule graph motifs via ring compression, and a $3$% improvement in SMILES validity over the DiGress model across all real-world molecule benchmarking datasets. Furthermore, our approach outperforms the state-of-the-art DeFoG method with respect to motif-conservation by roughly $4$%, as evidenced by high ChEMBL-likeness, QED and a newly introduced shingles distance score. The key ideas behind our approach are to use a combination of deterministic and random subgraph perturbations, so that the node and edge noise schedules are codependent; to modify the loss function of the training process in order to exploit the deterministic component of the schedule; and, to ``compress'' a collection of highly relevant carbon ring structures into supernodes in a way that allows for simple subsequent integration into the molecule scaffold.
|
Poster
|
AI Debate Aids Assessment of Controversial Claims
|
https://neurips.cc//virtual/2025/poster/117257
|
Salman Rahman, Issaka, Ashima Suvarna, Genglin Liu, James Shiffer, jaeyoung lee, Md Rizwan Parvez, Hamid Palangi, Shi Feng, Nanyun Peng, Yejin Choi, Julian Michael, Liwei Jiang, Saadia Gabriel
|
As AI grows more powerful, it will increasingly shape how we understand the world. But with this influence comes the risk of amplifying misinformation and deepening social divides—especially on consequential topics like public health where factual accuracy directly impacts well-being. Scalable Oversight aims to ensure AI truthfulness by enabling humans to supervise systems that may exceed human capabilities---yet humans themselves hold different beliefs and biases that impair their judgment. We study whether AI debate can guide biased judges toward the truth by having two AI systems debate opposing sides of controversial COVID-19 factuality claims where people hold strong prior beliefs. We conduct two studies: one with human judges holding either mainstream or skeptical beliefs evaluating factuality claims through AI-assisted debate or consultancy protocols, and a second examining the same problem with personalized AI judges designed to mimic these different human belief systems. In our human study, we find that debate—where two AI advisor systems present opposing evidence-based arguments—consistently improves judgment accuracy and confidence calibration, outperforming consultancy with a single-advisor system by 10\% overall. The improvement is most significant for judges with mainstream beliefs (+15.2\% accuracy), though debate also helps skeptical judges who initially misjudge claims move toward accurate views (+4.7\% accuracy). In our AI judge study, we find that AI judges with human-like personas achieve even higher accuracy (78.5\%) than human judges (70.1\%) and default AI judges without personas (69.8\%), suggesting their potential for supervising frontier AI models. These findings highlight AI debate as a promising path toward scalable, bias-resilient oversight---leveraging both diverse human and AI judgments to move closer to truth in contested domains.
|
Poster
|
AiDE-Q: Synthetic Labeled Datasets Can Enhance Learning Models for Quantum Property Estimation
|
https://neurips.cc//virtual/2025/poster/115703
|
Xinbiao Wang, Yuxuan Du, Zihan Lou, Yang Qian, Kaining Zhang, Yong Luo, Bo Du, Dacheng Tao
|
Quantum many-body problems are central to various scientific disciplines, yet their ground-state properties are intrinsically challenging to estimate. Recent advances in deep learning (DL) offer potential solutions in this field, complementing prior purely classical and quantum approaches. However, existing DL-based models typically assume access to a large-scale and noiseless labeled dataset collected by infinite sampling. This idealization raises fundamental concerns about their practical utility, especially given the limited availability of quantum hardware in the near term. To unleash the power of these DL-based models, we propose AiDE-Q (\underline{a}utomat\underline{i}c \underline{d}ata \underline{e}ngine for \underline{q}uantum property estimation), an effective framework that addresses this challenge by iteratively generating high-quality synthetic labeled datasets. Specifically, AiDE-Q utilizes a confidence-check method to assess the quality of synthetic labels and continuously improves the employed DL models with the identified high-quality synthetic dataset. To verify the effectiveness of AiDE-Q, we conduct extensive numerical simulations on a diverse set of quantum many-body and molecular systems, with up to 50 qubits. The results show that AiDE-Q enhances prediction performance for various reference learning models, with improvements of up to $14.2\\%$. Moreover, we exhibit that a basic supervised learning model integrated with AiDE-Q outperforms advanced reference models, highlighting the importance of a synthetic dataset. Our work paves the way for more efficient and practical applications of DL for quantum property estimation.
|
Poster
|
AI-Generated Video Detection via Perceptual Straightening
|
https://neurips.cc//virtual/2025/poster/118520
|
Christian Internò, Robert Geirhos, Markus Olhofer, Sunny Liu, Barbara Hammer, David Klindt
|
The rapid advancement of generative AI enables highly realistic synthetic video, posing significant challenges for content authentication and raising urgent concerns about misuse. Existing detection methods often struggle with generalization and capturing subtle temporal inconsistencies. We propose $ReStraV$ ($Re$presentation $Stra$ightening for $V$ideo), a novel approach to distinguish natural from AI-generated videos. Inspired by the ``perceptual straightening'' hypothesis—which suggests real-world video trajectories become more straight in neural representation domain—we analyze deviations from this expected geometric property. Using a pre-trained self-supervised vision transformer (DINOv2), we quantify the temporal curvature and stepwise distance in the model's representation domain. We aggregate statistical and signals descriptors of these measures for each video and train a classifier. Our analysis shows that AI-generated videos exhibit significantly different curvature and distance patterns compared to real videos. A lightweight classifier achieves state-of-the-art detection performance (e.g., $97.17$ % accuracy and $98.63$ % AUROC on the VidProM benchmark, substantially outperforming existing image- and video-based methods. ReStraV is computationally efficient, it is offering a low-cost and effective detection solution. This work provides new insights into using neural representation geometry for AI-generated video detection.
|
Poster
|
A Implies B: Circuit Analysis in LLMs for Propositional Logical Reasoning
|
https://neurips.cc//virtual/2025/poster/118508
|
Guan Zhe Hong, Nishanth Dikkala, Enming Luo, Cyrus Rashtchian, Xin Wang, Rina Panigrahy
|
Due to the size and complexity of modern large language models (LLMs), it has proven challenging to uncover the underlying mechanisms that models use to solve reasoning problems. For instance, is their reasoning for a specific problem localized to certain parts of the network? Do they break down the reasoning problem into modular components that are then executed as sequential steps as we go deeper in the model? To better understand the reasoning capability of LLMs, we study a minimal propositional logic problem that requires combining multiple facts to arrive at a solution. By studying this problem on Mistral and Gemma models, up to 27B parameters, we illuminate the core components the models use to solve such logic problems. From a mechanistic interpretability point of view, we use causal mediation analysis to uncover the pathways and components of the LLMs' reasoning processes. Then, we offer fine-grained insights into the functions of attention heads in different layers. We not only find a sparse circuit that computes the answer, but we decompose it into sub-circuits that have four distinct and modular uses. Finally, we reveal that three distinct models -- Mistral-7B, Gemma-2-9B and Gemma-2-27B -- contain analogous but not identical mechanisms.
|
Poster
|
AION-1: Omnimodal Foundation Model for Astronomical Sciences
|
https://neurips.cc//virtual/2025/poster/119776
|
Francois Lanusse, Liam Parker, Jeff Shen, Ollie Liu, Tom Hehir, Leopoldo Sarra, Lucas Meyer, Micah Bowles, Sebastian Wagner-Carena, Helen Qu, Siavash Golkar, Alberto Bietti, Hatim Bourfoune, Pierre Cornette, Keiya Hirashima, Geraud Krawezik, Ruben Ohana, Nicholas Lourie, Michael McCabe, Rudy Morel, Payel Mukhopadhyay, Mariel Pettee, Kyunghyun Cho, Miles Cranmer, Shirley Ho
|
While foundation models have shown promise across a variety of fields, astronomy lacks a unified framework for joint modeling across its highly diverse data modalities. In this paper, we present AION-1, the first large-scale multimodal foundation family of models for astronomy. AION-1 enables arbitrary transformations between heterogeneous data types using a two-stage architecture: modality-specific tokenization followed by transformer-based masked modeling of cross-modal token sequences. Trained on over 200M astronomical objects, AION-1 demonstrates strong performance across regression, classification, generation, and object retrieval tasks. Beyond astronomy, AION-1 provides a scalable blueprint for multimodal scientific foundation models that can seamlessly integrate heterogeneous combinations of real-world observations. Our model release is entirely open source, including the dataset, training script, and weights.
|
Poster
|
AI Research Agents for Machine Learning: Search, Exploration, and Generalization in MLE-bench
|
https://neurips.cc//virtual/2025/poster/117980
|
Edan Toledo, Karen Hambardzumyan, Martin Josifoski, RISHI HAZRA, Nicolas Baldwin, Alexis Audran-Reiss, Michael Kuchnik, Despoina Magka, Minqi Jiang, Alisia Lupidi, Andrei Lupu, Roberta Raileanu, Kelvin Niu, Tatiana Shavrina, Jean-Christophe Gagnon-Audet, Michael Shvartsman, Shagun Sodhani, Alexander Miller, Abhishek Charnalia, Derek Dunfield, Carole-Jean Wu, Pontus Lars Erik Saito Stenetorp, Nicola Cancedda, Jakob Foerster, Yoram Bachrach
|
AI research agents are demonstrating great potential to accelerate scientific progress by automating the design, implementation and training of machine learning models. We focus on methods for improving agents' performance on MLE-bench, a challenging benchmark where agents compete in Kaggle competitions to solve real-world machine learning problems. We formalize AI research agents as search policies that navigate a space of candidate solutions, iteratively modifying them using operators. By designing and systematically varying different operator sets and search policies (Greedy, MCTS, Evolutionary), we show that their interplay is critical for achieving high performance. Our best pairing of search strategy and operator set achieves a new state-of-the-art result on MLE-bench lite, increasing the success rate of achieving a Kaggle medal from 39.8 % to 47 %. Our investigation underscores the importance of jointly considering the search strategy, operator design, and evaluation methodology in advancing automated machine learning.
|
Poster
|
AI-Researcher: Autonomous Scientific Innovation
|
https://neurips.cc//virtual/2025/poster/116385
|
Jiabin Tang, Lianghao Xia, Zhonghang Li, Chao Huang
|
The powerful reasoning capabilities of Large Language Models (LLMs) in mathematics and coding, combined with their ability to automate complex tasks through agentic frameworks, present unprecedented opportunities for accelerating scientific innovation. In this paper, we introduce AI-Researcher, a fully autonomous research system that transforms how AI-driven scientific discovery is conducted and evaluated. Our framework seamlessly orchestrates the complete research pipeline--from literature review and hypothesis generation to algorithm implementation and publication-ready manuscript preparation--with minimal human intervention. To rigorously assess autonomous research capabilities, we develop Scientist-Bench, a comprehensive benchmark comprising state-of-the-art papers across diverse AI research domains, featuring both guided innovation and open-ended exploration tasks. Through extensive experiments, we demonstrate that AI-Researcher achieves remarkable implementation success rates and produces research papers that approach human-level quality. This work establishes new foundations for autonomous scientific innovation that can complement human researchers by systematically exploring solution spaces beyond cognitive limitations.
|
Poster
|
A is for Absorption: Studying Feature Splitting and Absorption in Sparse Autoencoders
|
https://neurips.cc//virtual/2025/poster/118058
|
David Chanin, James Wilken-Smith, Tomáš Dulka, Hardik Bhatnagar, Satvik Golechha, Joseph Bloom
|
Sparse Autoencoders (SAEs) aim to decompose the activation space of large language models (LLMs) into human-interpretable latent directions or features. As we increase the number of features in the SAE, hierarchical features tend to split into finer features (“math” may split into “algebra”, “geometry”, etc.), a phenomenon referred to as feature splitting. However, we show that sparse decomposition and splitting of hierarchical features is not robust. Specifically, we show that seemingly monosemantic features fail to fire where they should, and instead get “absorbed” into their children features. We coin this phenomenon feature absorption, and show that it is caused by optimizing for sparsity in SAEs whenever the underlying features form a hierarchy. We introduce a metric to detect absorption in SAEs, and validate our findings empirically on hundreds of LLM SAEs. Our investigation suggests that varying SAE sizes or sparsity is insufficient to solve this issue. We discuss the implications of feature absorption in SAEs and some potential approaches to solve the fundamental theoretical issues before SAEs can be used for interpreting LLMs robustly and at scale.
|
Poster
|
A kernel conditional two-sample test
|
https://neurips.cc//virtual/2025/poster/116663
|
Pierre-François Massiani, Christian Fiedler, Lukas Haverbeck, Friedrich Solowjow, Sebastian Trimpe
|
We propose a framework for hypothesis testing on conditional probability distributions, which we then use to construct conditional two-sample statistical tests. These tests identify the inputs --- called covariates in this context --- where two conditional expectations differ with high probability. Our key idea is to transform confidence bounds of a learning method into a conditional two-sample test, and we instantiate this principle for kernel ridge regression (KRR) and conditional kernel mean embeddings. We generalize existing pointwise-in-time or time-uniform confidence bounds for KRR to previously-inaccessible yet essential cases such as infinite-dimensional outputs with non-trace-class kernels. These bounds enable circumventing the need for independent data in our statistical tests, since they allow online sampling. We also introduce bootstrapping schemes leveraging the parametric form of testing thresholds identified in theory to avoid tuning inaccessible parameters, making our method readily applicable in practice. Such conditional two-sample tests are especially relevant in applications where data arrive sequentially or non-independently, or when output distributions vary with operational parameters. We demonstrate their utility through examples in process monitoring and comparison of dynamical systems. Overall, our results establish a comprehensive foundation for conditional two-sample testing, from theoretical guarantees to practical implementation, and advance the state-of-the-art on the concentration of vector-valued least squares estimation.
|
Poster
|
A Latent Multilayer Graphical Model For Complex, Interdependent Systems
|
https://neurips.cc//virtual/2025/poster/117984
|
Martin Ondrus, Ivor Cribben, Yang Feng
|
Networks have been extensively used in and provided novel insights in a wide variety of research areas. However, many real-world systems are in fact a ``network of networks'', or a multilayer network, which interact as components of a larger multimodal system. A major difficulty in this multilayer framework is the estimation of interlayer edges or connections. In this work, we propose a new estimation method, called multilayer sparse low-rank inverse covariance estimation (multiSLICE), which estimates the interlayer edges. multiSLICE bridges latent variable Gaussian graphical methods with multilayer networks, offering a flexible framework for modeling processes with irregular sampling and heterogeneous graph structures. We develop an efficient computational algorithm to compute this estimator and establish theoretical conditions for the recoverability of the joint space and how inter-layer interactions influence joint parameter estimation, and we provide theoretical bounds on their relationships. Finally, we rigorously evaluate our method on both simulated and multimodal neuroimaging data, demonstrating improvements over state-of-the-art approaches. All experiments are available in an Anonymous Github.
|
Poster
|
Alchemist: Turning Public Text-to-Image Data into Generative Gold
|
https://neurips.cc//virtual/2025/poster/121494
|
Valerii Startsev, Alexander Ustyuzhanin, Alexey Kirillov, Dmitry Baranchuk, Sergey Kastryulin
|
Pre-training equips text-to-image (T2I) models with broad world knowledge, but this alone is often insufficient to achieve high aesthetic quality and alignment. Consequently, supervised fine-tuning (SFT) is crucial for further refinement. However, its effectiveness highly depends on the quality of the fine-tuning dataset.Existing public SFT datasets frequently target narrow domains (e.g., anime or specific art styles), and the creation of high-quality, general-purpose SFT datasets remains a significant challenge.Current curation methods are often costly and struggle to identify truly impactful samples.This challenge is further complicated by the scarcity of public general-purpose datasets, as leading models often rely on large, proprietary, and poorly documented internal data, hindering broader research progress.This paper introduces a novel methodology for creating general-purpose SFT datasets by leveraging a pre-trained generative model as an estimator of high-impact training samples. We apply this methodology to construct and release Alchemist, a compact (3,350 samples) yet highly effective SFT dataset. Experiments demonstrate that Alchemist substantially improves the generative quality of five public T2I models while preserving diversity and style. Additionally, we release the fine-tuned models' weights to the public.
|
Poster
|
A learnability analysis on neuro-symbolic learning
|
https://neurips.cc//virtual/2025/poster/119006
|
Hao-Yuan He, Ming LI
|
This paper presents a comprehensive theoretical analysis of the learnability of neuro-symbolic (NeSy) tasks within hybrid systems. We characterize the learnability of NeSy tasks by their derived constraint satisfaction problems (DCSPs), demonstrating that a task is learnable if and only if its corresponding DCSP admits a unique solution. Under mild assumptions, we establish the sample complexity for learnable tasks and show that, for general tasks, the asymptotic expected concept error is controlled by the degree of disagreement among DCSP solutions. Our findings unify the characterization of learnability and the phenomenon of reasoning shortcuts, providing theoretical guarantees and actionable guidance for the principled design of NeSy systems.
|
Poster
|
A Learning-Augmented Approach to Online Allocation Problems
|
https://neurips.cc//virtual/2025/poster/120342
|
Ilan Cohen, Debmalya Panigrahi
|
In online allocation problems, an algorithm must choose from a set of options at each step, where each option incurs a set of costs/rewards associated with a set of $d$ agents. The goal is to minimize/maximize a function of the accumulated costs/rewards assigned to the agents over the course of the entire allocation process. Such problems are common in combinatorial optimization, including minimization problems such as machine scheduling and network routing, as well as maximization problems such as fair allocation for welfare maximization.In this paper, we develop a general learning-augmented algorithmic framework for online allocation problems that produces a nearly optimal solution using only a single $d$-dimensional vector of learned weights. Using this general framework, we derive learning-augmented online algorithms for a broad range of application problems in routing, scheduling, and fair allocation. Our main tool is convex programming duality, which may also have further implications for learning-augmented algorithms in the future.
|
Poster
|
A Learning-Augmented Exact Algorithm for Orienteering Problem with Time Windows
|
https://neurips.cc//virtual/2025/poster/118291
|
Guansheng Peng, Lining Xing, Fuyan Ma, Aldy Gunawan, Guopeng Song, Pieter Vansteenwegen
|
Recent years have witnessed a surge of interest in solving combinatorial optimization problems (COPs) using machine learning techniques. Motivated by this trend, we propose a learning-augmented exact approach for tackling an NP-hard COP, the Orienteering Problem with Time Windows, which aims to maximize the total score collected by visiting a subset of vertices in a graph within their time windows. Traditional exact algorithms rely heavily on domain expertise and meticulous design, making it hard to achieve further improvements. By leveraging deep learning models to learn effective relaxations of problem restrictions from data, our approach enables significant performance gains in an exact dynamic programming algorithm. We propose a novel graph convolutional network that predicts the directed edges defining the relaxation. The network is trained in a supervised manner, using optimal solutions as high-quality labels. Experimental results demonstrate that the proposed learning-augmented algorithm outperforms the state-of-the-art exact algorithm, achieving a 38% speedup on Solomon’s benchmark and more than a sevenfold improvement on the more challenging Cordeau’s benchmark.
|
Poster
|
ALE-Bench: A Benchmark for Long-Horizon Objective-Driven Algorithm Engineering
|
https://neurips.cc//virtual/2025/poster/121724
|
Yuki Imajuku, Kohki Horie, Yoichi Iwata, Kensho Aoki, Naohiro Takahashi, Takuya Akiba
|
How well do AI systems perform in algorithm engineering for hard optimization problems in domains such as package-delivery routing, crew scheduling, factory production planning, and power-grid balancing?We introduce $\textit{ALE-Bench}$, a new benchmark for evaluating AI systems on score-based algorithmic programming contests. Drawing on real tasks from the AtCoder Heuristic Contests, ALE-Bench presents optimization problems that are computationally hard and admit no known exact solution.Unlike short-duration, pass/fail coding benchmarks, ALE-Bench encourages iterative solution refinement over long time horizons.Our software framework supports interactive agent architectures that leverage test-run feedback and visualizations. Our evaluation of frontier LLMs revealed that while they demonstrate high performance on specific problems, a notable gap remains compared to humans in terms of consistency across problems and long-horizon problem-solving capabilities. This highlights the need for this benchmark to foster future AI advancements.
|
Poster
|
Algorithm- and Data-Dependent Generalization Bounds for Diffusion Models
|
https://neurips.cc//virtual/2025/poster/118406
|
Benjamin Dupuis, Dario Shariatian, Maxime Haddouche, Alain Durmus, Umut Simsekli
|
Score-based generative models (SGMs) have emerged as one of the most popular classes of generative models. A substantial body of work now exists on the analysis of SGMs, focusing either on discretization aspects or on their statistical performance. In the latter case, bounds have been derived, under various metrics, between the true data distribution and the distribution induced by the SGM, often demonstrating polynomial convergence rates with respect to the number of training samples. However, these approaches adopt a largely approximation theory viewpoint, which tends to be overly pessimistic and relatively coarse. In particular, they fail to fully explain the empirical success of SGMs or capture the role of the optimization algorithm used in practice to train the score network. To support this observation, we first present simple experiments illustrating the concrete impact of optimization hyperparameters on the generalization ability of the generated distribution. Then, this paper aims to bridge this theoretical gap by providing the first algorithmic- and data-dependent generalization analysis for SGMs. In particular, we establish bounds that explicitly account for the optimization dynamics of the learning algorithm, offering new insights into the generalization behavior of SGMs. Our theoretical findings are supported by empirical results on several datasets.
|
Poster
|
Algorithms and SQ Lower Bounds for Robustly Learning Real-valued Multi-Index Models
|
https://neurips.cc//virtual/2025/poster/117117
|
Ilias Diakonikolas, Giannis Iakovidis, Daniel Kane, Lisheng Ren
|
We study the complexity of learning real-valued Multi-Index Models (MIMs) under the Gaussian distribution. A $K$-MIM is a function $f:\mathbb{R}^d\to \mathbb{R}$ that depends only on the projection of its input onto a $K$-dimensional subspace. We give a general algorithm for PAC learning a broad class of MIMs with respect to the square loss, even in the presence of adversarial label noise. Moreover, we establish a nearly matching Statistical Query (SQ) lower bound, providing evidence that the complexity of our algorithm is qualitatively optimal as a function of the dimension. Specifically, we consider the class of bounded variation MIMs with the property that degree at most $m$ distinguishing moments exist with respect to projections onto any subspace. In the presence of adversarial label noise, the complexity of our learning algorithm is $d^{O(m)}2^{\mathrm{poly}(K/\epsilon)}$. For the realizable and independent noise settings, our algorithm incurs complexity $d^{O(m)}2^{\mathrm{poly}(K)}(1/\epsilon)^{O(K)}$. To complement our upper bound, we show that if for some subspace degree-$m$ distinguishing moments do not exist, then any SQ learner for the corresponding class of MIMs requires complexity $d^{\Omega(m)}$. As an application, we give the first efficient learner for the class of positive-homogeneous $L$-Lipschitz $K$-MIMs. The resulting algorithm has complexity $\mathrm{poly}(d) 2^{\mathrm{poly}(KL/\epsilon)}$. This gives a new PAC learning algorithm for Lipschitz homogeneous ReLU networks with complexity independent of the network size, removing the exponential dependence incurred in prior work.
|
Poster
|
AlgoTune: Can Language Models Speed Up General-Purpose Numerical Programs?
|
https://neurips.cc//virtual/2025/poster/121543
|
Ori Press, Brandon Amos, Haoyu Zhao, Yikai Wu, Samuel Ainsworth, Dominik Krupke, Patrick Kidger, Touqir Sajed, Bartolomeo Stellato, Jisun Park, Nathanael Bosch, Eli Meril, Albert Steppi, Arman Zharmagambetov, Fangzhao Zhang, David Pérez-Piñeiro, Alberto Mercurio, Ni Zhan, Talor Abramovich, Kilian Lieret, Hanlin Zhang, Shirley Huang, Matthias Bethge, Ofir Press
|
Despite progress in language model (LM) capabilities, evaluations have thus far focused on models' performance on tasks that humans have previously solved, including in programming (SWE-Bench) and mathematics (FrontierMath). We therefore propose testing models' ability to design and implement algorithms in an open-ended benchmark: We task LMs with writing code that efficiently solves computationally challenging problems in computer science, physics, and mathematics. Our AlgoTune benchmark consists of 120 tasks collected from domain experts and a framework for validating and timing LM-synthesized solution code, which is compared to reference implementations from popular open-source packages.In addition, we develop a baseline LM agent, AlgoTuner, and evaluate its performance across a suite of frontier models.AlgoTuner achieves an average 1.58x speedup against reference solvers, including methods from packages such as SciPy, scikit-learn and CVXPY.However, we find that current models fail to discover algorithmic innovations, instead preferring surface-level optimizations. We hope that AlgoTune catalyzes the development of LM agents exhibiting creative problem solving beyond state-of-the-art human performance.
|
Poster
|
Alias-Free ViT: Fractional Shift Invariance via Linear Attention
|
https://neurips.cc//virtual/2025/poster/118064
|
Hagay Michaeli, Daniel Soudry
|
Transformers have emerged as a competitive alternative to convnets in vision tasks, yet they lack the architectural inductive bias of convnets, which may hinder their potential performance. Specifically, Vision Transformers (ViTs) are not translation‑invariant and are more sensitive to minor image translations than standard convnets.Previous studies have shown, however, that convnets are also not perfectly shift‑invariant, due to aliasing in down‑sampling and non‑linear layers. Consequently, anti‑aliasing approaches have been proposed to certify convnets translation robustness. Building on this line of work, we propose an Alias‑Free ViT, which combines two main components. First, it uses alias-free down‑sampling and non‑linearities. Second, it uses linear cross‑covariance attention that is shift‑invariant to both integer and fractional translations.Our model maintains competitive performance in image classification and outperforms similar‑sized models in terms of robustness to adversarial translations.
|
Poster
|
AlignAb: Pareto-Optimal Energy Alignment for Designing Nature-Like Antibodies
|
https://neurips.cc//virtual/2025/poster/119755
|
Yibo Wen, Chenwei Xu, Jerry Yao-Chieh Hu, Kaize Ding, Han Liu
|
We present a three-stage framework for training deep learning models specializing in antibody sequence-structure co-design.We first pre-train a language model using millions of antibody sequence data.Then, we employ the learned representations to guide the training of a diffusion model for joint optimization over both sequence and structure of antibodies. During the final alignment stage, we optimize the model to favor antibodies with low repulsion and high attraction to the antigen binding site, enhancing the rationality and functionality of the designs.To mitigate conflicting energy preferences, we extend AbDPO (Antibody Direct Preference Optimization) to guide the model toward Pareto optimality under multiple energy-based alignment objectives. Furthermore, we adopt an iterative learning paradigm with temperature scaling, enabling the model to benefit from diverse online datasets without requiring additional data.In practice, our proposed methods achieve high stability and efficiency in producing a better Pareto front of antibody designs compared to top samples generated by baselines and previous alignment techniques.Through extensive experiments, we showcase the superior performance of our methods in generating nature-like antibodies with high binding affinity.
|
Poster
|
Align-DA: Align Score-based Atmospheric Data Assimilation with Multiple Preferences
|
https://neurips.cc//virtual/2025/poster/119125
|
Jing-An Sun, Hang Fan, Junchao Gong, Ben Fei, Kun Chen, Fenghua Ling, zhangwenlong, Wanghan Xu, Li Yan, Pierre Gentine, LEI BAI
|
Data assimilation (DA) aims to estimate the full state of a dynamical system by combining partial and noisy observations with a prior model forecast, commonly referred to as the background. In atmospheric applications, the problem is fundamentally ill-posed due to the sparsity of observations relative to the high-dimensional state space. Traditional methods address this challenge by simplifying background priors to regularize the solution, which are empirical and require continual tuning for application. Inspired by alignment techniques in text-to-image diffusion models, we propose Align-DA, which formulates DA as a generative process and uses reward signals to guide—replacing manual tuning with data-driven alignment. Specifically, we train a score-based model in the latent space to approximate the background-conditioned prior, and align it using three complementary reward signals for DA: (1) assimilation accuracy, (2) forecast skill initialized from the assimilated state, and (3) physical consistency of the analysis fields. Experiments with multiple reward signals demonstrate consistent improvements in analysis quality across different evaluation metrics and observation-guidance strategies. These results show that preference alignment, implemented as a soft constraint, can automatically adapt complex priors tailored to DA, offering a promising new direction for advancing the field.
|
Poster
|
AlignedGen: Aligning Style Across Generated Images
|
https://neurips.cc//virtual/2025/poster/117223
|
Jiexuan Zhang, Yiheng Du, Qian Wang, Weiqi Li, Yu Gu, Jian Zhang
|
Diffusion-based generative models struggle to maintain high style consistency across generated images via text description. Although several style-aligned image generation methods have been proposed to address this issue, they exhibit suboptimal performance and are primarily built upon the U-Net architecture, limiting their compatibility with MM-DiT diffusion models like Flux that has emerged as a predominant model in the field of image generation. To address these limitations, we propose $\textit{\textbf{AlignedGen}}$, a novel training-free style-aligned image generation method for Flux to significantly enhance style consistency across generated images. Specifically, AlignedGen incorporates two key components to achieve this: Shifted Position Embedding (ShiftPE) and Selective Shared Attention (SSA) layer. ShiftPE alleviates the text controllability degradation observed in prior methods when applied to Flux through its non-overlapping position indices design, while SSA further enhances style consistency across images. In addition, our method can be seamlessly integrated with various controllable generation technologies (e.g., subject-driven generation, depth control), demonstrating broad applicability across diverse scenarios. Extensive experimental results validate that our method effectively enhances style consistency across generated images while maintaining favorable text controllability.
|
Poster
|
Aligning by Misaligning: Boundary-aware Curriculum Learning for Multimodal Alignment
|
https://neurips.cc//virtual/2025/poster/118266
|
Hua Ye, Hang Ding, Siyuan Chen, Yiyang Jiang, changyuan zhang, Xuan Zhang
|
Most multimodal models treat every negative pair alike, ignoring the ambiguous negatives that differ from the positive by only a small detail. We propose Boundary-A ware Curriculum with Local Attention(BACL), a lightweight add-on that turns these borderline cases into a curriculum signal. A Boundary-aware Negative Sampler gradually raises difficulty, while a Contrastive Local Attention loss highlights where the mismatch occurs. The two modules are fully differentiable and work with any off-the-shelf dual encoder. Theory predicts a fast $\tilde{\mathcal{O}}(1/n)$ error rate; practice shows up to +32 \% R@1 over CLIP and new SOTA on four large-scale benchmarks, all without extra labels.
|
Poster
|
Aligning Compound AI Systems via System-level DPO
|
https://neurips.cc//virtual/2025/poster/119801
|
Xiangwen Wang, Yibo Jacky Zhang, Zhoujie Ding, Katherine Tsai, Haolun Wu, Sanmi Koyejo
|
Compound AI systems, comprising multiple interacting components such as LLMs, foundation models, and external tools, have demonstrated remarkable improvements compared to single models in various tasks. To ensure their effective deployment in real-world applications, aligning these systems with human preferences is crucial. However, aligning the compound system via policy optimization, unlike the alignment of a single model, is challenging for two main reasons: (i) non-differentiable interactions between components make end-to-end gradient-based optimization method inapplicable, and (ii) system-level preferences cannot be directly transformed into component-level preferences. To address these challenges, we first formulate compound AI systems as Directed Acyclic Graphs (DAGs), explicitly modeling both component interactions and the associated data flows. Building on this formulation, we introduce SysDPO, a framework that extends Direct Preference Optimization (DPO) to enable joint system-level alignment. We propose two variants, SysDPO-Direct and SysDPO-Sampling, tailored for scenarios depending on whether we construct a system-specific preference dataset. We empirically demonstrate the effectiveness of our approach across two applications: the joint alignment of a language model and a diffusion model, and the joint alignment of an LLM collaboration system.
|
Poster
|
Aligning Evaluation with Clinical Priorities: Calibration, Label Shift, and Error Costs
|
https://neurips.cc//virtual/2025/poster/117206
|
Gerardo Flores, Alyssa H. Smith, Julia Fukuyama, Ashia Wilson
|
Machine learning-based decision support systems are increasingly deployed in clinical settings, where probabilistic scoring functions are used to inform and prioritize patient management decisions.However, widely used scoring rules, such as accuracy and AUC-ROC, fail to adequately reflect key clinical priorities, including calibration, robustness to distributional shifts, and sensitivity to asymmetric error costs.In this work, we propose a principled yet practical evaluation framework for selecting calibrated thresholded classifiers that explicitly accounts for uncertainty in class prevalences and domain-specific cost asymmetries.Building on the theory of proper scoring rules, particularly the Schervish representation, we derive an adjusted variant of cross-entropy (log score) that averages cost-weighted performance over clinically relevant ranges of class balance.The resulting evaluation is simple to apply, sensitive to clinical deployment conditions, and designed to prioritize models that are both calibrated and robust to real-world variations.
|
Poster
|
Aligning Text-to-Image Diffusion Models to Human Preference by Classification
|
https://neurips.cc//virtual/2025/poster/117391
|
Longquan Dai, Xiaolu Wei, wang he, Shaomeng Wang, Jinhui Tang
|
Text-to-image diffusion models are typically trained on large-scale web data, often resulting in outputs that misalign with human preferences. Inspired by preference learning in large language models, we propose ABC (Alignment by Classification), a simple yet effective framework for aligning diffusion models with human preferences. In contrast to prior DPO-based methods that depend on suboptimal supervised fine-tuned (SFT) reference models, ABC assumes access to an ideal reference model perfectly aligned with human intent and reformulates alignment as a classification problem. Under this view, we recognize that preference data naturally forms a semi-supervised classification setting. To address this, we propose a data augmentation strategy that transforms preference comparisons into fully supervised training signals. We then introduce a classification-based ABC loss to guide alignment. Our alignment by classification approach could effectively steer the diffusion model toward the behavior of the ideal reference. Experiments on various diffusion models show that our ABC consistently outperforms existing baselines, offering a scalable and robust solution for preference-based text-to-image fine-tuning.
|
Poster
|
Aligning Text to Image in Diffusion Models is Easier Than You Think
|
https://neurips.cc//virtual/2025/poster/117814
|
Jaa-Yeon Lee, ByungHee Cha, Jeongsol Kim, Jong Chul Ye
|
While recent advancements in generative modeling have significantly improved text-image alignment, some residual misalignment between text and image representations still remains. Some approaches address this issue by fine-tuning models in terms of preference optimization, etc., which require tailored datasets. Orthogonal to these methods, we revisit the challenge from the perspective of representation alignment—an approach that has gained popularity with the success of REPresentation Alignment (REPA). We first argue that conventional text-to-image (T2I) diffusion models, typically trained on paired image and text data (i.e., positive pairs) by minimizing score matching or flow matching losses, is suboptimal from the standpoint of representation alignment. Instead, a better alignment can be achieved through contrastive learning that leverages existing dataset as both positive and negative pairs. To enable efficient alignment with pretrained models, we propose SoftREPA—a lightweight contrastive fine-tuning strategy that leverages soft text tokens for representation alignment. This approach improves alignment with minimal computational overhead by adding fewer than 1M trainable parameters to the pretrained model. Our theoretical analysis demonstrates that our method explicitly increases the mutual information between text and image representations, leading to enhanced semantic consistency. Experimental results across text-to-image generation and text-guided image editing tasks validate the effectiveness of our approach in improving the semantic consistency of T2I generative models.
|
Poster
|
Aligning Transformers with Continuous Feedback via Energy Rank Alignment
|
https://neurips.cc//virtual/2025/poster/117927
|
Shriram Chennakesavalu, Frank Hu, Sebastian Ibarraran, Grant Rotskoff
|
Searching through chemical space is an exceptionally challenging problem because the number of possible molecules grows combinatorially with the number of atoms. Large, autoregressive models trained on databases of chemical compounds have yielded powerful generators, but we still lack robust strategies for generating molecules with desired properties. This molecular search problem closely resembles the "alignment" problem for large language models, though for many chemical tasks we have a specific and easily evaluable reward function. Here, we introduce an algorithm called energy rank alignment (ERA) that leverages an explicit reward function to produce a gradient-based objective that we use to optimize autoregressive policies. We show theoretically that this algorithm is closely related to proximal policy optimization (PPO) and direct preference optimization (DPO), but has a minimizer that converges to an ideal Gibbs-Boltzmann distribution with the reward playing the role of an energy function. Furthermore, this algorithm is highly scalable, does not require reinforcement learning, and performs well relative to DPO when the number of preference observations per pairing is small. We deploy this approach to align molecular transformers and protein language models to generate molecules and protein sequences, respectively, with externally specifiedproperties and find that it does so robustly, searching through diverse parts of chemical space.
|
Poster
|
Aligning What Matters: Masked Latent Adaptation for Text-to-Audio-Video Generation
|
https://neurips.cc//virtual/2025/poster/118857
|
Jiyang Zheng, Siqi Pan, Yu Yao, Zhaoqing Wang, Dadong Wang, Tongliang Liu
|
Text-to-Audio-Video (T2AV) generation aims to produce temporally and semantically aligned visual and auditory content from natural language descriptions. While recent progress in text-to-audio and text-to-video models has improved generation quality within each modality, jointly modeling them remains challenging due to incomplete and asymmetric correspondence: audio often reflects only a subset of the visual scene, and vice versa. Naively enforcing full alignment introduces semantic noise and temporal mismatches. To address this, we propose a novel framework that performs selective cross-modal alignment through a learnable masking mechanism, enabling the model to isolate and align only the shared latent components relevant to both modalities. This mechanism is integrated into an adaptation module that interfaces with pretrained encoders and decoders from latent video and audio diffusion models, preserving their generative capacity with reduced training overhead. Theoretically, we show that our masked objective provably recovers the minimal set of shared latent variables across modalities. Empirically, our method achieves state-of-the-art performance on standard T2AV benchmarks, demonstrating significant improvements in audiovisual synchronization and semantic consistency.
|
Poster
|
Alignment of Large Language Models with Constrained Learning
|
https://neurips.cc//virtual/2025/poster/117670
|
Botong Zhang, Shuo Li, Ignacio Hounie, Osbert Bastani, Dongsheng Ding, Alejandro Ribeiro
|
We study the problem of computing an optimal large language model (LLM) policy for a constrained alignment problem, where the goal is to maximize a primary reward objective while satisfying constraints on secondary utilities. Despite the popularity of Lagrangian-based LLM policy search in constrained alignment, iterative primal-dual methods often fail to converge, and non-iterative dual-based methods do not achieve optimality in the LLM parameter space. To address these challenges, we employ Lagrangian duality to develop an iterative dual-based alignment method that alternates between updating the LLM policy via Lagrangian maximization and updating the dual variable via dual descent. In theory, we characterize the primal-dual gap between the primal value in the distribution space and the dual value in the LLM parameter space. We further quantify the optimality gap of the learned LLM policies at near-optimal dual variables with respect to both the objective and the constraint functions. These results prove that dual-based alignment methods can find an optimal constrained LLM policy, up to an LLM parametrization gap. We demonstrate the effectiveness and merits of our approach through extensive experiments conducted on the PKU-SafeRLHF dataset.
|
Poster
|
AlignVLM: Bridging Vision and Language Latent Spaces for Multimodal Document Understanding
|
https://neurips.cc//virtual/2025/poster/115435
|
Ahmed Masry, Juan Rodriguez, Tianyu Zhang, Suyuchen Wang, Chao Wang, Aarash Feizi, Akshay Kalkunte Suresh, Abhay Puri, Xiangru Jian, Pierre-André Noël, Sathwik Tejaswi Madhusudhan, Marco Pedersoli, Bang Liu, Nicolas Chapados, Yoshua Bengio, Enamul Hoque, Chris Pal, Issam Hadj Laradji, David Vazquez, Perouz Taslakian, Spandana Gella, Sai Rajeswar Mudumba
|
Aligning visual features with language embeddings is a key challenge in vision-language models (VLMs). The performance of such models hinges on having a good connector that maps visual features generated by a vision encoder to a shared embedding space with the LLM while preserving semantic similarity. Existing connectors, such as multilayer perceptrons (MLPs), often produce out-of-distribution or noisy inputs, leading to misalignment between the modalities. In this work, we propose a novel vision-text alignment method, AlignVLM, that maps visual features to a weighted average of LLM text embeddings. Our approach leverages the linguistic priors encoded by the LLM to ensure that visual features are mapped to regions of the space that the LLM can effectively interpret. AlignVLM is particularly effective for document understanding tasks, where scanned document images must be accurately mapped to their textual content. Our extensive experiments show that AlignVLM achieves state-of-the-art performance compared to prior alignment methods. We provide further analysis demonstrating improved vision-text feature alignment and robustness to noise.
|
Poster
|
Align Your Flow: Scaling Continuous-Time Flow Map Distillation
|
https://neurips.cc//virtual/2025/poster/115909
|
Amirmojtaba Sabour, Sanja Fidler, Karsten Kreis
|
Diffusion- and flow-based models have emerged as state-of-the-art generative modeling approaches, but they require many sampling steps. Consistency models can distill these models into efficient one-step generators; however, unlike flow- and diffusion-based methods, their performance inevitably degrades when increasing the number of steps, which we show both analytically and empirically.Flow maps generalize these approaches by connecting any two noise levels in a single step and remain effective across all step counts. In this paper, we introduce two new continuous-time objectives for training flow maps, along with additional novel training techniques, generalizing existing consistency and flow matching objectives. We further demonstrate that autoguidance can improve performance, using a low-quality model for guidance during distillation, and an additional boost can be achieved by adversarial finetuning, with minimal loss in sample diversity.We extensively validate our flow map models, called *Align Your Flow*, on challenging image generation benchmarks and achieve state-of-the-art few-step generation performance on both ImageNet 64x64 and 512x512, using small and efficient neural networks. Finally, we show text-to-image flow map models that outperform all existing non-adversarially trained few-step samplers in text-conditioned synthesis.
|
Poster
|
ALINE: Joint Amortization for Bayesian Inference and Active Data Acquisition
|
https://neurips.cc//virtual/2025/poster/117138
|
Daolang Huang, Xinyi Wen, Ayush Bharti, Samuel Kaski, Luigi Acerbi
|
Many critical applications, from autonomous scientific discovery to personalized medicine, demand systems that can both strategically acquire the most informative data and instantaneously perform inference based upon it. While amortized methods for Bayesian inference and experimental design offer part of the solution, neither approach is optimal in the most general and challenging task, where new data needs to be collected for instant inference. To tackle this issue, we introduce the Amortized Active Learning and Inference Engine (ALINE), a unified framework for amortized Bayesian inference and active data acquisition. ALINE leverages a transformer architecture trained via reinforcement learning with a reward based on self-estimated information gain provided by its own integrated inference component. This allows it to strategically query informative data points while simultaneously refining its predictions. Moreover, ALINE can selectively direct its querying strategy towards specific subsets of model parameters or designated predictive tasks, optimizing for posterior estimation, data prediction, or a mixture thereof. Empirical results on regression-based active learning, classical Bayesian experimental design benchmarks, and a psychometric model with selectively targeted parameters demonstrate that ALINE delivers both instant and accurate inference along with efficient selection of informative points.
|
Poster
|
AliO: Output Alignment Matters in Long-Term Time Series Forecasting
|
https://neurips.cc//virtual/2025/poster/119426
|
Kwangryeol Park, Jaeho Kim, Seulki Lee
|
Long-term Time Series Forecasting (LTSF) tasks, which leverage the current data sequence as input to predict the future sequence, have become increasingly crucial in real-world applications such as weather forecasting and planning of electricity consumption. However, state-of-the-art LTSF models often fail to achieve prediction output alignment for the same timestamps across lagged input sequences. Instead, these models exhibit low output alignment, resulting in fluctuation in prediction outputs for the same timestamps, undermining the model's reliability. To address this, we propose AliO (Align Outputs), a novel approach designed to improve the output alignment of LTSF models by reducing the discrepancies between prediction outputs for the same timestamps in both the time and frequency domains. To measure output alignment, we introduce a new metric, TAM (Time Alignment Metric), which quantifies the alignment between prediction outputs, whereas existing metrics such as MSE only capture the distance between prediction outputs and ground truths. Experimental results show that AliO effectively improves the output alignment, i.e., up to 58.2\% in TAM, while maintaining or enhancing the forecasting performance (up to 27.5\%). This improved output alignment increases the reliability of the LTSF models, making them more applicable in real-world scenarios. The code implementation is on an anonymous GitHub repository.
|
Poster
|
A Little Depth Goes a Long Way: The Expressive Power of Log-Depth Transformers
|
https://neurips.cc//virtual/2025/poster/119830
|
Will Merrill, Ashish Sabharwal
|
Recent theoretical results show transformers cannot express sequential reasoning problems over long input lengths, intuitively because their computational *depth* is bounded. However, prior work treats the depth as a constant, leaving it unclear to what degree bounded depth may suffice for solving problems over short inputs, or how increasing the transformer's depth affects its expressive power. We address these questions by analyzing the expressive power of transformers whose depth can grow minimally with context length $n$. We show even highly uniform transformers with depth $\Theta(\log n)$ can express two important problems: *recognizing regular languages*, which captures state tracking abilities and was known to be expressible only by a non-standard non-uniform model of transformers, and *graph connectivity*, which underlies multi-step reasoning. Notably, both of these problems cannot be expressed by fixed-depth transformers under standard complexity conjectures, demonstrating the expressivity benefit of growing depth. Moreover, our theory quantitatively predicts how depth must grow with input length to express these problems, showing that depth scaling is more efficient than scaling width or chain-of-thought steps. Empirically, our detailed experiments designed to bridge the expressivity vs. learnability gap reveal that our theoretical depth requirements for regular language recognition match the practical depth requirements for successfully training transformers remarkably well. Thus, our results clarify precisely how depth affects transformers' reasoning capabilities, providing potential practical insights for designing models that are better at sequential reasoning.
|
Poster
|
Alleviating Hallucinations in Large Language Models through Multi-Model Contrastive Decoding and Dynamic Hallucination Detection
|
https://neurips.cc//virtual/2025/poster/118154
|
Chenyu Zhu, Yefeng Liu, Hao Zhang, Aowen Wang, Yangxue, Guanhua Chen, Longyue Wang, Weihua Luo, Kaifu Zhang
|
Despite their outstanding performance in numerous applications, large language models (LLMs) remain prone to hallucinations, generating content inconsistent with their pretraining corpora. Currently, almost all contrastive decoding approaches alleviate hallucinations by introducing a model susceptible to hallucinations and appropriately widening the contrastive logits gap between hallucinatory tokens and target tokens. However, although existing contrastive decoding methods mitigate hallucinations, they lack enough confidence in the factual accuracy of the generated content. In this work, we propose Multi-Model Contrastive Decoding (MCD), which integrates a pretrained language model with an evil model and a truthful model for contrastive decoding. Intuitively, a token is assigned a high probability only when deemed potentially hallucinatory by the evil model while being considered factual by the truthful model. This decoding strategy significantly enhances the model’s confidence in its generated responses and reduces potential hallucinations. Furthermore, we introduce a dynamic hallucination detection mechanism that facilitates token-by-token identification of hallucinations during generation and a tree-based revision mechanism to diminish hallucinations further. Extensive experimental evaluations demonstrate that our MCD strategy effectively reduces hallucinations in LLMs and outperforms state-of-the-art methods across various benchmarks.
|
Poster
|
Alligat0R: Pre-Training through Covisibility Segmentation for Relative Camera Pose Regression
|
https://neurips.cc//virtual/2025/poster/115161
|
Thibaut Loiseau, Guillaume Bourmaud, Vincent Lepetit
|
Pre-training techniques have greatly advanced computer vision, with CroCo’s cross-view completion approach yielding impressive results in tasks like 3D reconstruction and pose regression. However, cross-view completion is ill-posed in non-covisible regions, limiting its effectiveness. We introduce Alligat0R, a novel pre-training approach that replaces cross-view learning with a covisibility segmentation task. Our method predicts whether each pixel in one image is covisible in the second image, occluded, or outside the field of view, making the pre-training effective in both covisible and non-covisible regions, and provides interpretable predictions. To support this, we present Cub3, a large-scale dataset with 5M image pairs and dense covisibility annotations derived from the nuScenes and ScanNet datasets. Cub3 includes diverse scenarios with varying degrees of overlap. The experiments show that our novel pre-training method Alligat0R significantly outperforms CroCo in relative pose regression. Alligat0R and Cub3 will be made publicly available.
|
Poster
|
All Proxy Rewards are Bad, Can We Hedge to Make Some Useful?
|
https://neurips.cc//virtual/2025/poster/116653
|
Hadi Khalaf, Claudio Mayrink Verdun, Alex Oesterling, Himabindu Lakkaraju, Flavio Calmon
|
A common paradigm to improve the performance of large language models is optimizing for a reward model. Reward models assign a numerical score to LLM outputs indicating, for example, which response would likely be preferred by a user or is most aligned with safety goals. However, reward models are never perfect. They inevitably function as proxies for complex desiderata such as correctness, helpfulness, and safety. By overoptimizing for a misspecified reward, we can subvert intended alignment goals and reduce overall performance - a phenomenon commonly referred to as reward hacking. In this work, we characterize reward hacking in inference-time alignment and demonstrate when and how we can mitigate it by hedging on the proxy reward. Hedging represents a tactical choice to avoid placing undue confidence in high but potentially misleading proxy reward signals. We study reward hacking under Best-of-$n$ (BoN) sampling, along with two inference-time methods, namely a novel method Best-of-Poisson (BoP) and Soft-Best-of-$n$ (SBoN), which introduce a parameter to control our confidence in the reward. We then propose $\texttt{HedgeTune}$ as an efficient algorithm to find the optimal hedging parameter. We demonstrate through experiments that hedging mitigates reward hacking and achieves superior distortion-reward tradeoffs with minimal computational overhead.
|
Poster
|
All that structure matches does not glitter
|
https://neurips.cc//virtual/2025/poster/121504
|
Maya Martirossyan, Thomas Egg, Philipp Höllmer, George Karypis, Mark Transtrum, Adrian Roitberg, Mingjie Liu, Richard Hennig, Ellad Tadmor, Stefano Martiniani
|
Generative models for materials, especially inorganic crystals, have attracted significant interest for their potential to transform the theoretical prediction of novel compounds and structures. Progress in this area depends critically on robust benchmarks and minimal, information-rich datasets that enable efficient and meaningful model evaluation. This paper critically examines commonly used datasets, methodologies, and evaluation metrics for the crystal structure prediction task—predicting the most likely structures given a chemical composition—and offers concrete solutions. We focus on three key issues: First, materials datasets should contain diverse and unique crystal structures; for example, we show that the widely-utilized carbon-24 dataset only contains $\approx 40$% unique structures, with duplicates differing only by the choice of unit cell representation. Second, materials datasets need to be split with care, rather than randomly, if polymorphs of many different compositions are numerous—which we discover to be the case for the perov-5 dataset. Third, benchmarks for evaluation of generative models can be misleading if used uncritically—for example, the reporting of a 'match rate' metric without consideration of the structural complexity that can be exhibited by identical building blocks (atoms). To address these oft-overlooked issues, we introduce several fixes. We provide revised versions of the carbon-24 dataset: one with duplicates removed, one deduplicated and split by number of atoms $N$, and two containing only duplicates. We also propose a new split for the perov-5 dataset that ensures polymorphs are grouped within the same training, validation, or test set in order to set a more sensible standard for benchmarking model performance. Finally, we present METRe and cRMSE, new model evaluation metrics that can correctly handle materials datasets with polymorphs.
|
Poster
|
All You Need is One: Capsule Prompt Tuning with a Single Vector
|
https://neurips.cc//virtual/2025/poster/117588
|
Yiyang Liu, James Liang, Heng Fan, Wenhao Yang, Yiming Cui, Xiaotian Han, Lifu Huangg, Dongfang Liu, Qifan Wang, Cheng Han
|
Prompt-based learning has emerged as a parameter-efficient finetuning (PEFT) approach to facilitate Large Language Model (LLM) adaptation to downstream tasks by conditioning generation with task-aware guidance. Despite its successes, current prompt-based learning methods heavily rely on laborious grid searching for optimal prompt length and typically require considerable number of prompts, introducing additional computational burden. Worse yet, our pioneer findings indicate that the task-aware prompt design is inherently limited by its absence of instance-aware information, leading to a subtle attention interplay with the input sequence. In contrast, simply incorporating instance-aware information as a part of the guidance can enhance the prompt-tuned model performance without additional fine-tuning. Moreover, we find an interesting phenomenon, namely "attention anchor", that incorporating instance-aware tokens at the earliest position of the sequence can successfully preserve strong attention to critical structural information and exhibit more active attention interaction with all input tokens. In light of our observation, we introduce Capsule Prompt-Tuning (CaPT), an efficient and effective solution that leverages off-the-shelf, informative instance semantics into prompt-based learning. Our approach innovatively integrates both instance-aware and task-aware information in a nearly parameter-free manner (i.e., one single capsule prompt).Empirical results demonstrate that our method can exhibit superior performance across various language tasks (e.g., 84.03\% average accuracy on T5-Large), serving as an "attention anchor," while enjoying high parameter efficiency (e.g., 0.003\% of model parameters on Llama3.2-1B).
|
Poster
|
ALMGuard: Safety Shortcuts and Where to Find Them as Guardrails for Audio–Language Models
|
https://neurips.cc//virtual/2025/poster/115978
|
Weifei Jin, Yuxin Cao, Junjie Su, Minhui Xue, Jie Hao, Ke Xu, Jin Song Dong, Derui Wang
|
Recent advances in Audio-Language Models (ALMs) have significantly improved multimodal understanding capabilities. However, the introduction of the audio modality also brings new and unique vulnerability vectors. Previous studies have proposed jailbreak attacks that specifically target ALMs, revealing that defenses directly transferred from traditional audio adversarial attacks or text-based LLM jailbreaks are largely ineffective against these ALM-specific threats.To address this issue, we propose ALMGuard, the first defense framework tailored to ALMs. Based on the assumption that safety-aligned shortcuts naturally exist in ALMs, we design a method to identify universal Shortcut Activation Perturbations (SAPs) that serve as triggers that activate the safety shortcuts to safeguard ALMs at inference time. To better sift out effective triggers while preserving the model’s availability on benign tasks, we further propose Mel-Gradient Sparse Mask (M-GSM), which restricts perturbations to Mel-frequency bins that are sensitive to jailbreaks but insensitive to speech understanding. Both theoretical analyses and empirical results demonstrate the robustness of our method against both seen and unseen attacks. ALMGuard reduces the average success rate of the most advanced ALM-specific jailbreak attacks to 4.6% across four models, establishing it as the new state-of-the-art in the field. Furthermore, evaluations on benign benchmarks confirm that our method does not cause a significant degradation in model availability.
|
Poster
|
AlphaBeta is not as good as you think: a new probabilistic model to better analyze deterministic game-solving algorithms
|
https://neurips.cc//virtual/2025/poster/115177
|
Raphael Boige, Amine Boumaza, Bruno Scherrer
|
Deterministic game-solving algorithms are conventionally analyzed in the light of their average-case complexity against a distribution of random game-trees, where leaf values are independently sampled from a fixed distribution. This simplified model enables uncluttered mathematical analysis, revealing two key properties: root value distributions asymptotically collapse to a single fixed value for finite-valued trees, and all reasonable algorithms achieve global optimality. However, these findings are artifacts of the model’s design—its long criticized independence assumption strips games of structural complexity, producing trivial instances where no algorithm faces meaningful challenges. To address this limitation, we introduce a new probabilistic model that incrementally constructs game-trees using a fixed level-wise conditional distribution. By enforcing ancestor dependency, a critical structural feature of real-world games, our framework generates problems with adjustable difficulty while retaining some form of analytical tractability. For several algorithms, including AlphaBeta and Scout, we derive recursive formulas characterizing their average-case complexities under this model. These allow us to rigorously compare algorithms on deep game-trees, where Monte-Carlo simulations are no longer feasible. While asymptotically, all algorithms seem to converge to identical branching factor (a result analogous to those of independence-based models), deep finite trees reveal stark differences: AlphaBeta incurs a significantly larger constant multiplicative factor compared to algorithms like Scout, leading to a substantial practical slowdown. Our framework sheds new light on classical game-solving algorithms, offering rigorous evidence and analytical tools to advance the understanding of these methods under a more realistic, challenging, and yet tractable model.
|
Poster
|
AlphaDecay: Module-wise Weight Decay for Heavy-Tailed Balancing in LLMs
|
https://neurips.cc//virtual/2025/poster/118480
|
Di He, Ajay Jaiswal, Songjun Tu, Li Shen, Ganzhao Yuan, Shiwei Liu, Lu Yin
|
Weight decay is a standard regularization technique for training large language models (LLMs). While it is common to assign a uniform decay rate to every layer, this approach overlooks the structural diversity of LLMs and the varying spectral properties across modules. In this paper, we introduce AlphaDecay, a simple yet effective method that adaptively assigns different weight decay strengths to each module of an LLM. Our approach is guided by Heavy-Tailed Self-Regularization (HT-SR) theory, which analyzes the empirical spectral density (ESD) of weight correlation matrices to quantify “heavy-tailedness.” Modules exhibiting more pronounced heavy-tailed ESDs, reflecting stronger feature learning, are assigned weaker decay, while modules with lighter-tailed spectra receive stronger decay. Our method leverages tailored weight decay assignments to balance the module-wise differences in spectral properties, leading to improved performance. Extensive pre-training tasks with various model sizes from 60M to 1B demonstrate that {\method} achieves better perplexity and generalization than conventional uniform decay and other adaptive decay baselines.
|
Poster
|
AlphaFold Database Debiasing for Robust Inverse Folding
|
https://neurips.cc//virtual/2025/poster/119664
|
Cheng Tan, Zhenxiao Cao, Zhangyang Gao, Siyuan Li, Yufei Huang, Stan Z. Li
|
The AlphaFold Protein Structure Database (AFDB) offers unparalleled structural coverage at near-experimental accuracy, positioning it as a valuable resource for data-driven protein design. However, its direct use in training deep models that are sensitive to fine-grained atomic geometry—such as inverse folding—exposes a critical limitation. Comparative analysis of structural feature distributions reveals that AFDB structures exhibit distinct statistical regularities, reflecting a systematic geometric bias that deviates from the conformational diversity found in experimentally determined structures from the Protein Data Bank (PDB). While AFDB structures are cleaner and more idealized, PDB structures capture the intrinsic variability and physical realism essential for generalization in downstream tasks. To address this discrepancy, we introduce a Debiasing Structure AutoEncoder (DeSAE) that learns to reconstruct native-like conformations from intentionally corrupted backbone geometries. By training the model to recover plausible structural states, DeSAE implicitly captures a more robust and natural structural manifold. At inference, applying DeSAE to AFDB structures produces debiased structures that significantly improve inverse folding performance across multiple benchmarks. This work highlights the critical impact of subtle systematic biases in predicted structures and presents a principled framework for debiasing, significantly boosting the performance of structure-based learning tasks like inverse folding.
|
Poster
|
AlphaZero Neural Scaling and Zipf's Law: a Tale of Board Games and Power Laws
|
https://neurips.cc//virtual/2025/poster/118808
|
Oren Neumann, Claudius Gros
|
Neural scaling laws are observed in a range of domains, to date with no clear understanding of why they occur. Recent theories suggest that loss power laws arise from Zipf's law, a power law observed in domains like natural language. One theory suggests that language scaling laws emerge when Zipf-distributed task quanta are learned in descending order of frequency. In this paper we examine power-law scaling in AlphaZero, a reinforcement learning algorithm, using a theory of language-model scaling. We find that game states in training and inference data scale with Zipf's law, which is known to arise from the tree structure of the environment, and examine the correlation between scaling-law and Zipf's-law exponents. In agreement with quanta scaling theory, we find that agents optimize state loss in descending order of frequency, even though this order scales inversely with modelling complexity. We also find that inverse scaling, the failure of models to improve with size, is correlated with unusual Zipf curves where end-game states are among the most frequent states. We show evidence that larger models shift their focus to these less-important states, sacrificing their understanding of important early-game states.
|
Poster
|
ALTER: All-in-One Layer Pruning and Temporal Expert Routing for Efficient Diffusion Generation
|
https://neurips.cc//virtual/2025/poster/120357
|
Xiaomeng Yang, LEI LU, Qihui Fan, Changdi Yang, Juyi Lin, Yanzhi Wang, Xuan Zhang, Shangqian Gao
|
Diffusion models have demonstrated exceptional capabilities in generating high-fidelity images. However, their iterative denoising process results in significant computational overhead during inference, limiting their practical deployment in resource-constrained environments. Existing acceleration methods often adopt uniform strategies that fail to capture the temporal variations during diffusion generation, while the commonly adopted sequential $\textit{pruning-then-fine-tuning strategy}$ suffers from sub-optimality due to the misalignment between pruning decisions made on pretrained weights and the model’s final parameters. To address these limitations, we introduce $\textbf{ALTER}$: $\textbf{A}$ll-in-One $\textbf{L}$ayer Pruning and $\textbf{T}$emporal $\textbf{E}$xpoert $\textbf{R}$outing, a unified framework that transforms diffusion models into a mixture of efficient temporal experts.ALTER achieves a single-stage optimization that unifies layer pruning, expert routing, and model fine-tuning by employing a trainable hypernetwork, which dynamically generates layer pruning decisions and manages timestep routing to specialized, pruned expert sub-networks throughout the ongoing fine-tuning of the UNet. This unified co-optimization strategy enables significant efficiency gains while preserving high generative quality. Specifically, ALTER achieves same-level visual fidelity to the original 50-step Stable Diffusion v2.1 model while utilizing only 25.9\% of its total MACs with just 20 inference steps and delivering a 3.64$\times$ speedup through 35\% sparsity.
|
Poster
|
Alternating Gradient Flows: A Theory of Feature Learning in Two-layer Neural Networks
|
https://neurips.cc//virtual/2025/poster/115627
|
Daniel Kunin, Giovanni Luca Marchetti, Feng Chen, Dhruva Karkada, James Simon, Michael Deweese, Surya Ganguli, Nina Miolane
|
What features neural networks learn, and how, remains an open question. In this paper, we introduce Alternating Gradient Flows (AGF), an algorithmic framework that describes the dynamics of feature learning in two-layer networks trained from small initialization. Prior works have shown that gradient flow in this regime exhibits a staircase-like loss curve, alternating between plateaus where neurons slowly align to useful directions and sharp drops where neurons rapidly grow in norm. AGF approximates this behavior as an alternating two-step process: maximizing a utility function over dormant neurons and minimizing a cost function over active ones. AGF begins with all neurons dormant. At each round, a dormant neuron activates, triggering the acquisition of a feature, and a drop in the loss. AGF quantifies the order, timing, and magnitude of these drops, matching experiments across architectures. We show that AGF unifies and extends existing saddle-to-saddle analyses in fully connected linear networks and attention-only linear transformers, where the learned features are singular modes and principal components, respectively. In diagonal linear networks, we prove AGF converges to gradient flow in the limit of vanishing initialization. Applying AGF to quadratic networks trained to perform modular addition, we give the first complete characterization of the training dynamics, revealing that networks learn Fourier features in decreasing order of coefficient magnitude. Altogether, AGF offers a promising step towards understanding feature learning in neural networks.
|
Poster
|
AltLoRA: Towards Better Gradient Approximation in Low-Rank Adaptation with Alternating Projections
|
https://neurips.cc//virtual/2025/poster/119533
|
Xin Yu, Yujia Wang, Jinghui Chen, Lingzhou Xue
|
Low-Rank Adaptation (LoRA) has emerged as an effective technique for reducing memory overhead in fine-tuning large language models. However, it often suffers from sub-optimal performance compared with full fine-tuning since the update is constrained in the low-rank space. Recent variants such as LoRA-Pro attempt to mitigate this by adjusting the gradients of the low-rank matrices to approximate the full gradient. However, LoRA-Pro's solution is not unique, and different solutions can lead to significantly varying performance in ablation studies. Besides, to incorporate momentum or adaptive optimization design, approaches like LoRA-Pro must first compute the equivalent gradient, causing a higher memory cost close to full fine-tuning. A key challenge remains in integrating momentum properly into the low-rank space with lower memory cost. In this work, we propose AltLoRA, an alternating projection method that avoids the difficulties in gradient approximation brought by the joint update design, meanwhile integrating momentum without higher memory complexity. Our theoretical analysis provides convergence guarantees and further shows that AltLoRA enables stable feature learning and robustness to transformation invariance. Extensive experiments across multiple tasks demonstrate that AltLoRA outperforms LoRA and its variants, narrowing the gap toward full fine-tuning while preserving superior memory efficiency.
|
Poster
|
ALTo: Adaptive-Length Tokenizer for Autoregressive Mask Generation
|
https://neurips.cc//virtual/2025/poster/116479
|
Lingfeng Wang, Hualing Lin, Senda Chen, Tao Wang, Changxu Cheng, Yangyang Zhong, Dong Zheng, Wuyue Zhao
|
While humans effortlessly draw visual objects and shapes by adaptively allocating attention based on their complexity, existing multimodal large language models (MLLMs) remain constrained by rigid token representations. Bridging this gap, we propose ALTo, an adaptive length tokenizer for autoregressive mask generation. To achieve this, a novel token length predictor is designed, along with a length regularization term and a differentiable token chunking strategy. We further build ALToLLM that seamlessly integrates ALTo into MLLM. Preferences on the trade-offs between mask quality and efficiency is implemented by group relative policy optimization (GRPO). Experiments demonstrate that ALToLLM achieves state-of-the-art performance with adaptive token cost on popular segmentation benchmarks. Code and models will be released.
|
Poster
|
A machine learning approach that beats large Rubik's cubes
|
https://neurips.cc//virtual/2025/poster/120075
|
Alexander Chervov, Kirill Khoruzhii, Nikita Bukhal, Jalal Naghiyev, Vladislav Zamkovoy, Ivan Koltsov, Lyudmila Cheldieva, Arsenii Sychev, Arsenii Lenin, Mark Obozov, Egor Urvanov, Alexey Romanov
|
The paper proposes a novel machine learning-based approach to the pathfinding problem on extremely large graphs. This method leverages diffusion distance estimation via a neural network and uses beam search for pathfinding. We demonstrate its efficiency by finding solutions for 4x4x4 and 5x5x5 Rubik's cubes with unprecedentedly short solution lengths, outperforming all available solvers and introducing the first machine learning solver beyond the 3x3x3 case. In particular, it surpasses every single case of the combined best results in the Kaggle Santa 2023 challenge, which involved over 1,000 teams. For the 3x3x3 Rubik's cube, our approach achieves an optimality rate exceeding 98%, matching the performance of task-specific solvers and significantly outperforming prior solutions such as DeepCubeA (60.3%) and EfficientCube (69.6%). Our solution in its current implementation is approximately 25.6 times faster in solving 3x3x3 Rubik's cubes while requiring up to 8.5 times less model training time than the most efficient state-of-the-art competitor. Finally, it is demonstrated that even a single agent trained using a relatively small number of examples can robustly solve a broad range of puzzles represented by Cayley graphs of size up to $10^{145}$, confirming the generality of the proposed method.
|
Poster
|
A Markov Decision Process for Variable Selection in Branch & Bound
|
https://neurips.cc//virtual/2025/poster/120351
|
Paul STRANG, Zacharie ALES, Côme Bissuel, Safia Kedad-Sidhoum, Olivier JUAN, Emmanuel Rachelson
|
Mixed-Integer Linear Programming (MILP) is a powerful framework used to address a wide range of NP-hard combinatorial optimization problems, often solved by Branch and bound (B&B). A key factor influencing the performance of B&B solvers is the variable selection heuristic governing branching decisions. Recent contributions have sought to adapt reinforcement learning (RL) algorithms to the B&B setting to learn optimal branching policies, through Markov Decision Processes (MDP) inspired formulations, and ad hoc convergence theorems and algorithms. In this work, we introduce B&B MDPs, a principled vanilla MDP formulation for variable selection in B&B, allowing to leverage a broad range of RL algorithms for the purpose of learning optimal B&B heuristics. Computational experiments validate our model empirically, as our branching agent outperforms prior state-of-the-art RL agents on four standard MILP benchmarks.
|
Poster
|
AMBER: Adaptive Mesh Generation by Iterative Mesh Resolution Prediction
|
https://neurips.cc//virtual/2025/poster/119306
|
Niklas Freymuth, Tobias Würth, Nicolas Schreiber, Balázs Gyenes, Andreas Boltres, Johannes Mitsch, Aleksandar Taranovic, Tai Hoang, Philipp Dahlinger, Philipp Becker, Luise Kärger, Gerhard Neumann
|
The cost and accuracy of simulating complex physical systems using the Finite Element Method (FEM) scales with the resolution of the underlying mesh. Adaptive meshes improve computational efficiency by refining resolution in critical regions, but typically require task-specific heuristics or cumbersome manual design by a human expert. We propose Adaptive Meshing By Expert Reconstruction (AMBER), a supervised learning approach to mesh adaptation. Starting from a coarse mesh, AMBER iteratively predicts the sizing field, i.e., a function mapping from the geometry to the local element size of the target mesh, and uses this prediction to produce a new intermediate mesh using an out-of-the-box mesh generator. This process is enabled through a hierarchical graph neural network, and relies on data augmentation by automatically projecting expert labels onto AMBER-generated data during training. We evaluate AMBER on 2D and 3D datasets, including classical physics problems, mechanical components, and real-world industrial designs with human expert meshes. AMBER generalizes to unseen geometries and consistently outperforms multiple recent baselines, including ones using Graph and Convolutional Neural Networks, and Reinforcement Learning-based approaches.
|
Poster
|
Ambient Diffusion Omni: Training Good Models with Bad Data
|
https://neurips.cc//virtual/2025/poster/118464
|
Giannis Daras, Adrian Rodriguez-Munoz, Adam Klivans, Antonio Torralba, Constantinos Daskalakis
|
We show how to use low-quality, synthetic, and out-of-distribution images to improve the quality of a diffusion model. Typically, diffusion models are trained on curated datasets that emerge from highly filtered data pools from the Web and other sources. We show that there is immense value in the lower-quality images that are often discarded. We present Ambient Diffusion Omni, a simple, principled framework to train diffusion models that can extract signal from arbitrarily images during training. Our framework exploits two properties of natural images -- spectral power law decay and locality. We first validate our framework by successfully training diffusion models with images synthetically corrupted by Gaussian blur, JPEG compression, and motion blur. We use our framework to achieve state-of-the-art ImageNet FID and we show significant improvements in both image quality and diversity for text-to-image generative modeling. The core insight is that noise dampens the initial skew between the desired high-quality distribution and the mixed distribution we actually observe. We provide rigorous theoretical justification for our approach by analyzing the trade-off between learning from biased data versus limited unbiased data across diffusion times.
|
Poster
|
Ambient Proteins - Training Diffusion Models on Noisy Structures
|
https://neurips.cc//virtual/2025/poster/117481
|
Giannis Daras, Jeffrey Ouyang-Zhang, Krithika Ravishankar, Constantinos Daskalakis, Adam Klivans, Daniel Diaz
|
We present Ambient Protein Diffusion, a framework for training protein diffusion models that generates structures with unprecedented diversity and quality. State-of-the-art generative models are trained on computationally derived structures from AlphaFold2 (AF), as experimentally determined structures are relatively scarce. The resulting models are therefore limited by the quality of synthetic datasets. Since the accuracy of AF predictions degrades with increasing protein length and complexity, de novo generation of long, complex proteins remains challenging. Ambient Protein Diffusion overcomes this problem by treating low-confidence AF structures as corrupted data. Rather than simply filtering out low-quality AF structures, our method adjusts the diffusion objective for each structure based on its corruption level, allowing the model to learn from both high and low quality structures. Empirically, ambient protein diffusion yields major improvements: on proteins with 700 residues, diversity increases from 45% to 85% from the previous state-of-the-art, and designability improves from 70% to 88%.
|
Poster
|
A-Mem: Agentic Memory for LLM Agents
|
https://neurips.cc//virtual/2025/poster/119020
|
Wujiang Xu, Kai Mei, Hang Gao, Juntao Tan, Zujie Liang, Yongfeng Zhang
|
While large language model (LLM) agents can effectively use external tools for complex real-world tasks, they require memory systems to leverage historical experiences. Current memory systems enable basic storage and retrieval but lack sophisticated memory organization, despite recent attempts to incorporate graph databases. Moreover, these systems' fixed operations and structures limit their adaptability across diverse tasks. To address this limitation, this paper proposes a novel agentic memory system for LLM agents that can dynamically organize memories in an agentic way. Following the basic principles of the Zettelkasten method, we designed our memory system to create interconnected knowledge networks through dynamic indexing and linking. When a new memory is added, we generate a comprehensive note containing multiple structured attributes, including contextual descriptions, keywords, and tags. The system then analyzes historical memories to identify relevant connections, establishing links where meaningful similarities exist. Additionally, this process enables memory evolution -- as new memories are integrated, they can trigger updates to the contextual representations and attributes of existing historical memories, allowing the memory network to continuously refine its understanding. Our approach combines the structured organization principles of Zettelkasten with the flexibility of agent-driven decision making, allowing for more adaptive and context-aware memory management.Empirical experiments on six foundation models show superior improvement against existing SOTA baselines. The code is available at \url{https://anonymous.4open.science/r/AgenticMemory-76B4}.
|
Poster
|
A Minimalist Example of Edge-of-Stability and Progressive Sharpening
|
https://neurips.cc//virtual/2025/poster/115100
|
Liming Liu, Zixuan Zhang, Simon Du, Tuo Zhao
|
Recent advances in deep learning optimization have unveiled two intriguing phenomena under large learning rates: Edge of Stability (EoS) and Progressive Sharpening (PS), challenging classical Gradient Descent (GD) analyses. Current research approaches, using either generalist frameworks or minimalist examples, face significant limitations in explaining these phenomena. This paper advances the minimalist approach by introducing a two-layer network with a two-dimensional input, where one dimension is relevant to the response and the other is irrelevant. Through this model, we rigorously prove the existence of progressive sharpening and self-stabilization under large learning rates, and establish non-asymptotic analysis of the training dynamics and sharpness along the entire GD trajectory. Besides, we connect our minimalist example to existing works by reconciling the existence of a well-behaved "stable set" between minimalist and generalist analyses, and extending the analysis of Gradient Flow Solution sharpness to our two-dimensional input scenario. These findings provide new insights into the EoS phenomenon from both parameter and input data distribution perspectives, potentially informing more effective optimization strategies in deep learning practice.
|
Poster
|
A Minimalistic Unified Framework for Incremental Learning across Image Restoration Tasks
|
https://neurips.cc//virtual/2025/poster/118487
|
Xiaoxuan Gong, Jie Ma
|
Existing research in low-level vision has shifted its focus from "one-by-one" task-specific methods to "all-in-one" multi-task unified architectures. However, current all-in-one image restoration approaches primarily aim to improve overall performance across a limited number of tasks. In contrast, how to incrementally add new image restoration capabilities on top of an existing model — that is, task-incremental learning — has been largely unexplored. To fill this research gap, we propose a minimalistic and universal paradigm for task-incremental learning called MINI. It addresses the problem of parameter interference across different tasks through a simple yet effective mechanism, enabling nearly forgetting-free task-incremental learning. Specifically, we design a special meta-convolution called MINI-Conv, which generates parameters solely through lightweight embeddings instead of complex convolutional networks or MLPs. This not only significantly reduces the number of parameters and computational overhead but also achieves complete parameter isolation across different tasks. Moreover, MINI-Conv can be seamlessly integrated as a plug-and-play replacement for any convolutional layer within existing backbone networks, endowing them with incremental learning capabilities. Therefore, our method is highly generalizable. Finally, we demonstrate that our method achieves state-of-the-art performance compared to existing incremental learning approaches across five common image restoration tasks. Moreover, the near forgetting-free nature of our method makes it highly competitive even against all-in-one image restoration methods trained in a full-supervised manner. Our code is available at https://github.com.
|
Poster
|
Among Us: A Sandbox for Measuring and Detecting Agentic Deception
|
https://neurips.cc//virtual/2025/poster/117514
|
Satvik Golechha, Adrià Garriga-Alonso
|
Prior studies on deception in language-based AI agents typically assess whether the agent produces a false statement about a topic, or makes a binary choice prompted by a goal, rather than allowing open-ended deceptive behavior to emerge in pursuit of a longer-term goal.To fix this, we introduce $\textit{Among Us}$, a sandbox social deception game where LLM-agents exhibit long-term, open-ended deception as a consequence of the game objectives. While most benchmarks saturate quickly, $\textit{Among Us}$ can be expected to last much longer, because it is a multi-player game far from equilibrium.Using the sandbox, we evaluate $18$ proprietary and open-weight LLMs and uncover a general trend:models trained with RL are comparatively much better at producing deception than detecting it.We evaluate the effectiveness of methods to detect lying and deception: logistic regression on the activations and sparse autoencoders (SAEs). We find that probes trained on a dataset of ``pretend you're a dishonest model: $\dots$'' generalize extremely well out-of-distribution, consistently obtaining AUROCs over 95% even when evaluated just on the deceptive statement, without the chain of thought. We also find two SAE features that work well at deception detection but are unable to steer the model to lie less.We hope our open-sourced sandbox, game logs, and probes serve to anticipate and mitigate deceptive behavior and capabilities in language-based agents.
|
Poster
|
AmorLIP: Efficient Language-Image Pretraining via Amortization
|
https://neurips.cc//virtual/2025/poster/118950
|
Haotian Sun, Yitong Li, Yuchen Zhuang, Niao He, Hanjun Dai, Bo Dai
|
Contrastive Language-Image Pretraining (CLIP) has demonstrated strong zero-shot performance across diverse downstream text-image tasks. Existing CLIP methods typically optimize a contrastive objective using negative samples drawn from each minibatch. To achieve robust representation learning, these methods require extremely large batch sizes and escalate computational demands to hundreds or even thousands of GPUs. Prior approaches to mitigate this issue often compromise downstream performance, prolong training duration, or face scalability challenges with very large datasets. To overcome these limitations, we propose AmorLIP, an efficient CLIP pretraining framework that amortizes expensive computations involved in contrastive learning through lightweight neural networks, which substantially improves training efficiency and performance. Leveraging insights from a spectral factorization of energy-based models, we introduce novel amortization objectives along with practical techniques to improve training stability. Extensive experiments across 38 downstream tasks demonstrate the superior zero-shot classification and retrieval capabilities of AmorLIP, consistently outperforming standard CLIP baselines with substantial relative improvements of up to 12.24%.
|
Poster
|
Amortized Active Generation of Pareto Sets
|
https://neurips.cc//virtual/2025/poster/116473
|
Daniel Steinberg, Asiri Wijesinghe, Rafael Oliveira, Piotr Koniusz, Cheng Soon Ong, Edwin Bonilla
|
We propose a new framework called active generation of Pareto sets (A-GPS) for online discrete black-box multi objective optimization (MOO) that learns a generative model of the Pareto set and supports a-posteriori preference conditioning. Our method actively learns a generative model conditioned on high-performance regions (active generation) using amortized variational inference. It uses a class probability estimator (CPE) for predicting Pareto-optimality and conditioning the generative model. Furthermore, motivated by discrete/mixed design problems where we must balance multiple competing objectives, it introduces preference direction vectors to capture subjective trade-offs. Thus, at each iteration, we update a generative model conditioned on Pareto set membership _and_ alignment with preference directions. Our method yields high-quality Pareto set approximations using only simple CPE guidance, avoids hyper-volume computation, and supports sampling at arbitrary trade-off points without retraining. Empirical results on synthetic functions and protein design benchmarks demonstrate strong sample efficiency and effective incorporation of users' preferences.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.