13 papers
Towards Understanding On-Policy Distillation through the Lens of Test-Time Scaling
Xinmu Ge, Zizhuo Zhang, Yu Huang +9
On-policy distillation (OPD) has emerged as a promising post-training technique for enhancing LLM reasoning. It is commonly believed to enable the student model to distill knowledg…
DICE: Entropy-Regularized Equilibrium Selection for Stable Multi-Agent LLM Coordination
Yi Xie, Zhanke Zhou, Chentao Cao +2
Multi-agent large language model (LLM) systems often fail to reliably outperform a single strong model equipped with best-of-N sampling. We argue that a core source of this instabi…
Beyond Homophily: Towards Generalized Graph Reconstruction Attack and Defense
Zhanke Zhou, Bo Han, Xuan Li +3
Graph neural networks (GNNs) are widely deployed on relational data, yet they can leak sensitive or proprietary information about the training graph adjacency, e.g., social ties, t…
The Easy, the Hard, and the Learnable: Confidence and Difficulty-Adaptive Policy Optimization for LLM Reasoning
Zhanke Zhou, Xiangyu Lu, Chentao Cao +4
RL with verifiable rewards can substantially improve LLM reasoning, yet standard GRPO-style training often treats easy, hard, and learnable questions alike through uniform sampling…
RewardFlow: Topology-Aware Reward Propagation on State Graphs for Agentic RL with Large Language Models
Xiao Feng, Bo Han, Zhanke Zhou +5
Reinforcement learning (RL) shows promise for enhancing LLM agentic reasoning, yet sparse terminal rewards hinder fine-grained optimization. Process reward modeling offers an alter…
Towards Understanding Valuable Preference Data for Large Language Model Alignment
Zizhuo Zhang, Qizhou Wang, Shanshan Ye +4
Large language model (LLM) alignment is typically achieved through learning from human preference comparisons, making the quality of preference data critical to its success. Existi…