collaborators

8 papers

cs.LG2026

Staleness-Learning Rate Scaling Laws for Asynchronous RLHF

Jingwei Song, Haofeng Xu, Jie Xiao +8

High-throughput RLHF systems often decouple rollout generation from policy optimization, leading to the use of stale rollouts during learner updates. In this work, we study the eff…

cs.LG2026

GAC: Stabilizing Asynchronous RL Training for LLMs via Gradient Alignment Control

Haofeng Xu, Junwei Su, Yukun Tian +3

Asynchronous execution is essential for scaling reinforcement learning (RL) to modern large model workloads, including large language models and AI agents, but it can fundamentally…

cs.LG2026

Step-TP: A Grounded, Step-Level Dataset with Chain-of-Thought Reasoning for LLM-Guided Tensor Program Optimization

Mengfan Liu, Da Zheng, Junwei Su +1

Despite the strong reasoning capabilities of large language models (LLMs), optimizing the execution efficiency of tensor programs remains challenging due to the need for precise, c…

cs.LG2026

Full-Graph vs. Mini-Batch Training: Comprehensive Analysis from a Batch Size and Fan-Out Size Perspective

Mengfan Liu, Da Zheng, Junwei Su +1

Full-graph and mini-batch Graph Neural Network (GNN) training approaches have distinct system design demands, making it crucial to choose the appropriate approach to develop. A cor…

cs.MA2026

GeomHerd: A Forward-looking Herding Quantification via Ricci Flow Geometry on Agent Interactive Simulations

Lake Yang, Junwei Su, Jingfeng Zeng +5

Herding -- where agents align their behaviors and act collectively -- is a central driver of market fragility and systemic risk. Existing approaches to quantify herding rely on pri…

cs.LG2025

BG-HGNN: Toward Efficient Learning for Complex Heterogeneous Graphs

Junwei Su, Lingjun Mao, Zheng Da +1

Heterogeneous graphs, comprising diverse node and edge types connected through varied relations, are ubiquitous in real-world applications. Message-passing heterogeneous graph neur…