CogDual: Enhancing Dual Cognition of LLMs via Reinforcement Learning with Implicit Rule-Based Rewards
[arXiv] / [GitHub Code]
I received my master's degree from Northeastern University under the supervision of Prof. Feiliang Ren. I am currently a Research Engineer at Qiantang Credit.
My research focuses on large language model agents, tool learning, and reinforcement learning, with a particular interest in building agents that can reason, interact with tools, and improve through experience. Previously, I held research and engineering internships at Meituan AI Search, Tencent AI Lab & Hunyuan Digital Human, and DeepLang AI.
Email: lyfei1126@gmail.com
CogDual: Enhancing Dual Cognition of LLMs via Reinforcement Learning with Implicit Rule-Based Rewards
[arXiv] / [GitHub Code]
CodeTool: Enhancing Programmatic Tool Invocation of LLMs via Process Supervision
[arXiv] / [GitHub Code]