collaborators

15 papers

cs.LG2026

VRPO: Rethinking Value Modeling for Robust RL under Noisy Supervision in LLM Post-Training

Dingwei Zhu, Shihan Dou, Zhiheng Xi +16

Reinforcement Learning (RL) in real-world environments often suffers from ambiguous or incomplete reward supervision, which undermines policy stability and generalization. Such noi…

cs.AI2026

Entropy Is Not Enough: Unlocking Effective Reinforcement Learning for Visual Reasoning via Vision-Anchored Token Selection

Senjie Jin, Peixin Wang, Boyang Liu +8

While token-level entropy is commonly recognized as effective for credit assignment in text-only reinforcement learning with verifiable rewards (RLVR), it remains unclear whether t…

cs.LG2026

DVPO: Distributional Value Modeling-based Policy Optimization for LLM Post-Training

Dingwei Zhu, Zhiheng Xi, Shihan Dou +15

Reinforcement learning (RL) has shown strong performance in LLM post-training, but real-world deployment often involves noisy or incomplete supervision. In such settings, complex a…

cs.CL2026

Locate, Steer, and Improve: A Practical Survey of Actionable Mechanistic Interpretability in Large Language Models

Hengyuan Zhang, Zhihao Zhang, Mingyang Wang +26

Mechanistic Interpretability (MI) has emerged as a vital approach to demystify the opaque decision-making of Large Language Models (LLMs). However, existing reviews primarily treat…

cs.AI2026

JFTA-Bench: Evaluate LLM's Ability of Tracking and Analyzing Malfunctions Using Fault Trees

Yuhui Wang, Zhixiong Yang, Ming Zhang +10

In the maintenance of complex systems, fault trees are used to locate problems and provide targeted solutions. To enable fault trees stored as images to be directly processed by la…

cs.CL2026

Why Reinforcement Fine-Tuning Enables MLLMs Preserve Prior Knowledge Better: A Data Perspective

Zhihao Zhang, Qiaole Dong, Qi Zhang +12

Post-training algorithms such as Supervised Fine-Tuning (SFT) and Reinforcement Fine-Tuning (RFT) are widely used to adapt (multimodal) large language models to downstream tasks. W…