Unsupervised Model Improvement Through Internal Coherence Maximization

EDINET-Bench: LLMs on Japanese Financial Tasks

AutoThink: Efficient LLM Reasoning with Adaptive Budgeting

System Prompt Learning for LLM Problem-Solving Strategies

OpenEvolve: Open Source AlphaEvolve Implementation

PTS: Pivotal Token Search

CameraBench: Understanding Video Motion

Step1X-Edit: General Image Editing Framework

VisuLogic: A Benchmark for Evaluating Visual Reasoning in Multi-modal Large Language Models

Does Reinforcement Learning Really Incentivize Reasoning Capacity in LLMs Beyond the Base Model?

Learning to Reason under Off-Policy Guidance

AI's Potential to Transform the World

Contents On the Nature of Time

MovieGen: A Detailed Review of Meta's Text-to-Video Generation System