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

Feb 3

Submitted by

akhaliq

s1: Simple test-time scaling

·
10 authors

Submitted by

yuhuixu

Reward-Guided Speculative Decoding for Efficient LLM Reasoning

·
8 authors

4

Submitted by

akhaliq

MatAnyone: Stable Video Matting with Consistent Memory Propagation

·
5 authors

Submitted by

Qika

Self-supervised Quantized Representation for Seamlessly Integrating Knowledge Graphs with Large Language Models

·
8 authors

4

Submitted by

SSamDav

Scalable-Softmax Is Superior for Attention

·
1 authors

Submitted by

wenhu

PixelWorld: Towards Perceiving Everything as Pixels

·
3 authors

2

Submitted by

nielsr

DINO-WM: World Models on Pre-trained Visual Features enable Zero-shot Planning

·
4 authors

Submitted by

akhaliq

Constitutional Classifiers: Defending against Universal Jailbreaks across Thousands of Hours of Red Teaming

·
43 authors

Submitted by

bcywinski

SAeUron: Interpretable Concept Unlearning in Diffusion Models with Sparse Autoencoders

·
2 authors

Submitted by

fabian-sp

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training

·
5 authors

Submitted by

mirshad7

Zero-Shot Novel View and Depth Synthesis with Multi-View Geometric Diffusion

·
5 authors

Submitted by

akhaliq

Trading Inference-Time Compute for Adversarial Robustness

·
11 authors

Submitted by

odabashi

Unraveling the Capabilities of Language Models in News Summarization

·
2 authors

Submitted by

nielsr

Fast Encoder-Based 3D from Casual Videos via Point Track Processing

·
3 authors

Submitted by

lwpyh

INT: Instance-Specific Negative Mining for Task-Generic Promptable Segmentation

·
3 authors

2

Submitted by

Dominic789654

ChunkKV: Semantic-Preserving KV Cache Compression for Efficient Long-Context LLM Inference

·
7 authors

2