5 papers
EG-GRPO: Evidence-Guided Step-level Credit Assignment for Deep Search Agents
Sijia Luo, Mengjie Gao, Yuxiao He +4
LLM-based search agents demonstrate strong capabilities in long-horizon information seeking and reasoning, typically trained via reinforcement learning. However, existing approache…
Selm: Self-Evolving Library Management for Multi-Skill Agent Optimization
Mengxi Hu, Yuxuan Hu, Jinbo Su +6
Large language model (LLM) agents increasingly use skills as a lightweight way to acquire task-specific knowledge. Recent work enables skills to evolve from task trajectories. Howe…
AsyncTLS: Efficient Generative LLM Inference with Asynchronous Two-level Sparse Attention
Yuxuan Hu, Jianchao Tan, Jiaqi Zhang +7
Long-context inference in LLMs faces the dual challenges of quadratic attention complexity and prohibitive KV cache memory. While token-level sparse attention offers superior accur…
Sparse-RL: Breaking the Memory Wall in LLM Reinforcement Learning via Stable Sparse Rollouts
Sijia Luo, Xiaokang Zhang, Yuxuan Hu +6
Reinforcement Learning (RL) has become essential for eliciting complex reasoning capabilities in Large Language Models (LLMs). However, the substantial memory overhead of storing K…
Optimizing Native Sparse Attention with Latent Attention and Local Global Alternating Strategies
Yuxuan Hu, Jianchao Tan, Jiaqi Zhang +7
In this work, we conduct a systematic analysis of Native Sparse Attention (NSA) and propose targeted improvements that enhance long-context modeling. A key insight is that alternat…