Zhiwei He

DeepMath-103K: A Large-Scale, Challenging, Decontaminated, and Verifiable Mathematical Dataset for Advancing Reasoning

RLVER: Reinforcement Learning with Verifiable Emotion Rewards for Empathetic Agents

DeepCompress: A Dual Reward Strategy for Dynamically Exploring and Compressing Reasoning Chains

DeepTheorem: Advancing LLM Reasoning for Theorem Proving Through Natural Language and Reinforcement Learning

Thoughts Are All Over the Place: On the Underthinking of Long Reasoning Models

Weak-to-Strong Preference Optimization: Stealing Reward from Weak Aligned Model

RaSA: Rank-Sharing Low-Rank Adaptation

Trust, But Verify: A Self-Verification Approach to Reinforcement Learning with Verifiable Rewards

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

The First Few Tokens Are All You Need: An Efficient and Effective Unsupervised Prefix Fine-Tuning Method for Reasoning Models

Do NOT Think That Much for 2+3=? On the Overthinking of Long Reasoning Models

Thoughts Are All Over the Place: On the Underthinking of o1-Like LLMs