new

Get trending papers in your email inbox once a day!

Get trending papers in your email inbox!

Daily Papers

byAK and the research community

May 21

Submitted by

Andy1621

Emerging Properties in Unified Multimodal Pretraining

·
12 authors

4

Submitted by

jt-zhang

SageAttention3: Microscaling FP4 Attention for Inference and An Exploration of 8-Bit Training

·
9 authors

Submitted by

unilm

Reward Reasoning Model

·
7 authors

2

Submitted by

QPHutu

Optimizing Anytime Reasoning via Budget Relative Policy Optimization

·
6 authors

Submitted by

HEmile

Neurosymbolic Diffusion Models

·
4 authors

Submitted by

TianheWu

VisualQuality-R1: Reasoning-Induced Image Quality Assessment via Reinforcement Learning to Rank

·
5 authors

Submitted by

myownskyW7

Visual Agentic Reinforcement Fine-Tuning

·
9 authors

Submitted by

FengTing

Latent Flow Transformer

·
6 authors

Submitted by

dariog

The Aloe Family Recipe for Open and Specialized Healthcare LLMs

·
13 authors

Submitted by

MrLight

General-Reasoner: Advancing LLM Reasoning Across All Domains

UWaterloo

University of Waterloo

Submitted by

YUSHUIWX

Think Only When You Need with Large Hybrid-Reasoning Models

·
10 authors

2

Submitted by

DKYoon

Reasoning Models Better Express Their Confidence

·
9 authors

Submitted by

wren93

VideoEval-Pro: Robust and Realistic Long Video Understanding Evaluation

·
7 authors

Submitted by

jiwonsong

Reasoning Path Compression: Compressing Generation Trajectories for Efficient LLM Reasoning

snu

Seoul National University

Submitted by

kaiyangzhou

Visionary-R1: Mitigating Shortcuts in Visual Reasoning with Reinforcement Learning

·
5 authors

Submitted by

akhaliq

Hunyuan-Game: Industrial-grade Intelligent Game Creation Model

·
50 authors

Submitted by

gpx333

Exploring Federated Pruning for Large Language Models

·
7 authors

Submitted by

SkAndMl

CS-Sum: A Benchmark for Code-Switching Dialogue Summarization and the Limits of Large Language Models

·
4 authors

3

Submitted by

kaiyangzhou

Training-Free Watermarking for Autoregressive Image Generation

HKBU

Hong Kong Baptist University

Submitted by

changdae

Visual Instruction Bottleneck Tuning

·
4 authors

2

Submitted by

kaiyangzhou

Fine-tuning Quantized Neural Networks with Zeroth-order Optimization

·
5 authors

Submitted by

huangsiteng

SSR: Enhancing Depth Perception in Vision-Language Models via Rationale-Guided Spatial Reasoning

·
8 authors

Submitted by

Ningyu

Two Experts Are All You Need for Steering Thinking: Reinforcing Cognitive Effort in MoE Reasoning Models Without Additional Training

·
15 authors

2

Submitted by

KID-22

NExT-Search: Rebuilding User Feedback Ecosystem for Generative AI Search

·
7 authors

2

Submitted by

Emperorizzis

Not All Correct Answers Are Equal: Why Your Distillation Source Matters

·
8 authors

2

Submitted by

bcywinski

Towards eliciting latent knowledge from LLMs with mechanistic interpretability

·
4 authors

Submitted by

tiantiaf

Vox-Profile: A Speech Foundation Model Benchmark for Characterizing Diverse Speaker and Speech Traits

·
12 authors

Submitted by

iliashum

Lessons from Defending Gemini Against Indirect Prompt Injections

·
14 authors

2

Submitted by

MaksimSTW

The Hallucination Tax of Reinforcement Finetuning

·
3 authors

2

Submitted by

Acatsama

Truth Neurons

·
5 authors

Submitted by

safal312

Warm Up Before You Train: Unlocking General Reasoning in Resource-Constrained Settings

·
5 authors

Submitted by

pierlj

Phare: A Safety Probe for Large Language Models

·
4 authors

Submitted by

iliashum

Fixing 7,400 Bugs for 1$: Cheap Crash-Site Program Repair

·
5 authors

2

Submitted by

sliuxl

MIGRATION-BENCH: Repository-Level Code Migration Benchmark from Java 8

·
11 authors

Submitted by

charleslipku

CoIn: Counting the Invisible Reasoning Tokens in Commercial Opaque LLM APIs

·
10 authors

Submitted by

DavidNguyen

CompeteSMoE -- Statistically Guaranteed Mixture of Experts Training via Competition

·
6 authors

Submitted by

Jianyuan1

Solve-Detect-Verify: Inference-Time Scaling with Flexible Generative Verifier

·
6 authors

2

Submitted by

hmarkc

Rethinking Optimal Verification Granularity for Compute-Efficient Test-Time Scaling

·
6 authors

Submitted by

KomeijiForce

Bidirectional LMs are Better Knowledge Memorizers? A Benchmark for Real-world Knowledge Injection

·
8 authors

Submitted by

Veeru

Understanding Gen Alpha Digital Language: Evaluation of LLM Safety Systems for Content Moderation

·
2 authors

3

Submitted by

kellycyy

Will AI Tell Lies to Save Sick Children? Litmus-Testing AI Values Prioritization with AIRiskDilemmas

·
7 authors

Submitted by

himel7

To Bias or Not to Bias: Detecting bias in News with bias-detector

·
3 authors

Submitted by

ChaoHuangCS

Learning to Highlight Audio by Watching Movies

·
8 authors

Submitted by

xianghe

Incorporating brain-inspired mechanisms for multimodal learning in artificial intelligence

·
6 authors

Submitted by

Wyattz23

Tokenization Constraints in LLMs: A Study of Symbolic and Arithmetic Reasoning Limits

·
5 authors

2

Submitted by

Jia-py

GeoRanker: Distance-Aware Ranking for Worldwide Image Geolocalization

·
5 authors

Submitted by

hwy9855

Masking in Multi-hop QA: An Analysis of How Language Models Perform with Context Permutation

·
4 authors

Submitted by

mohbattharani

KERL: Knowledge-Enhanced Personalized Recipe Recommendation using Large Language Models

·
2 authors

Submitted by

Marl

Dynadiff: Single-stage Decoding of Images from Continuously Evolving fMRI

·
3 authors

Submitted by

Manishemi

Void in Language Models

·
1 authors

Submitted by

florin-hf

The Distracting Effect: Understanding Irrelevant Passages in RAG

·
4 authors

2

Submitted by

jwgcurrie

Towards Embodied Cognition in Robots via Spatially Grounded Synthetic Worlds

·
7 authors

2

Submitted by

Beegbrain

Object-Centric Representations Improve Policy Generalization in Robot Manipulation

·
4 authors

2