14 papers
Evolving Together, Failing Apart: The Skill-Policy Co-evolution Trap in Self-Improving Agents
Yujie Feng, Xiaobin Wang, Jinluan Yang +5
Skill-policy co-evolution, in which an agent's policy model and its skill bank improve each other during training, is a promising paradigm for self-improving agents. Current evalua…
SkillScaffold: Scaffolding Agentic Reinforcement Learning with Policy-Adaptive Skill Credit Assignment and Selective Injection
Wenlin Zhong, Hao Wang, Kun Yuan +6
Agentic reinforcement learning (RL) enables large language models (LLMs) to learn from multi-turn interactions with verifiable outcomes. Textual skills can improve exploration by d…
TopoCurate:Modeling Interaction Topology for Tool-Use Agent Training
Jinluan Yang, Yuxin Liu, Zhengyu Chen +7
Training tool-use agents typically relies on outcome-based filtering: Supervised Fine-Tuning (SFT) on successful trajectories and Reinforcement Learning (RL) on pass-rate-selected…
ExpertWeaver: Unlocking the Inherent MoE in Dense LLMs with GLU Activation Patterns
Ziyu Zhao, Tong Zhu, Zhi Zhang +6
Mixture-of-Experts (MoE) effectively scales model capacity while preserving computational efficiency through sparse expert activation. However, training high-quality MoEs from scra…
SIGHT: Reinforcement Learning with Self-Evidence and Information-Gain Diverse Branching for Search Agent
Wenlin Zhong, Jinluan Yang, Yiquan Wu +3
Reinforcement Learning (RL) has empowered Large Language Models (LLMs) to master autonomous search for complex question answering. However, particularly within multi-turn search sc…
LongCat-Flash-Thinking-2601 Technical Report
Meituan LongCat Team, Anchun Gui, Bei Li +162
We introduce LongCat-Flash-Thinking-2601, a 560-billion-parameter open-source Mixture-of-Experts (MoE) reasoning model with superior agentic reasoning capability. LongCat-Flash-Thi…
Scaling and Transferability of Annealing Strategies in Large Language Model Training
Siqi Wang, Zhengyu Chen, Teng Xiao +5
Learning rate scheduling is crucial for training large language models, yet understanding the optimal annealing strategies across different model configurations remains challenging…
Reinforcement Learning for Tool-Integrated Interleaved Thinking towards Cross-Domain Generalization
Zhengyu Chen, Jinluan Yang, Teng Xiao +6
Recent advances in large language models (LLMs) have demonstrated remarkable capabilities in reasoning and tool utilization. However, the generalization of tool-augmented reinforce…
Unifying Adversarial Perturbation for Graph Neural Networks
Jinluan Yang, Ruihao Zhang, Zhengyu Chen +2
This paper studies the vulnerability of Graph Neural Networks (GNNs) to adversarial attacks on node features and graph structure. Various methods have implemented adversarial train…
Ada-R1: Hybrid-CoT via Bi-Level Adaptive Reasoning Optimization
Haotian Luo, Haiying He, Yibo Wang +6
Recently, long-thought reasoning models achieve strong performance on complex reasoning tasks, but often incur substantial inference overhead, making efficiency a critical concern.…
Mix Data or Merge Models? Balancing the Helpfulness, Honesty, and Harmlessness of Large Language Model via Model Merging
Jinluan Yang, Dingnan Jin, Anke Tang +10
Achieving balanced alignment of large language models (LLMs) in terms of Helpfulness, Honesty, and Harmlessness (3H optimization) constitutes a cornerstone of responsible AI. Exist…
Mitigating the Backdoor Effect for Multi-Task Model Merging via Safety-Aware Subspace
Jinluan Yang, Anke Tang, Didi Zhu +3
Model merging has gained significant attention as a cost-effective approach to integrate multiple single-task fine-tuned models into a unified one that can perform well on multiple…
Leveraging Invariant Principle for Heterophilic Graph Structure Distribution Shifts
Jinluan Yang, Zhengyu Chen, Teng Xiao +3
Heterophilic Graph Neural Networks (HGNNs) have shown promising results for semi-supervised learning tasks on graphs. Notably, most real-world heterophilic graphs are composed of a…
Learning to Reweight for Graph Neural Network
Zhengyu Chen, Teng Xiao, Kun Kuang +6
Graph Neural Networks (GNNs) show promising results for graph tasks. However, existing GNNs' generalization ability will degrade when there exist distribution shifts between testin…