AI 연구 논문 데일리

s1: Simple test-time scaling

Reward-Guided Speculative Decoding for Efficient LLM Reasoning

MatAnyone: Stable Video Matting with Consistent Memory Propagation

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

Scalable-Softmax Is Superior for Attention

PixelWorld: Towards Perceiving Everything as Pixels

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

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

SAeUron: Interpretable Concept Unlearning in Diffusion Models with Sparse Autoencoders

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

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

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

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

Unraveling the Capabilities of Language Models in News Summarization

Trading Inference-Time Compute for Adversarial Robustness

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