Paper introduces SePO for self-evolving system prompt optimization

Arxiv CS.AI (top papers)

SePO automates system prompt optimization for agents without modifying underlying models, using a prompt agent that learns from past experiences. The method achieves improvements across diverse tasks while keeping instructions human-readable and model-agnostic.

prompt-engineering

See 11 more sources

AIP: A Graph Representation for Learning and Governing Agent Skills7 days agoZachary Blumenfeld, Jim Webber

RUBAS: Rubric-Based Reinforcement Learning for Agent Safety6 days agoXian Qi Loye, Qinglin Su, Zhexin Zhang, Shiyao Cui, Qi Zhu, Fei Mi, Hongning Wang, Minlie Huang

SePO: Self-Evolving Prompt Agent for System Prompt Optimization7 days agoWangcheng Tao, Han Wu, Weng-Fai Wong

DuMate-DeepResearch: An Auditable Multi-Agent System with Recursive Search and Rubric-Grounded Reasoning2 days agoLingyong Yan, Can Xu, Yukun Zhao, Wenxuan Li, Qingyang Chen, Jiulong Wu, Wenli Song, Xiangnan Li, Weixian Shi, Yiqun Chen, Xuchen Ma, Yuchen Li, Jiashu Zhao, Shuaiqiang Wang, Jianmin Wu, Dawei Yin

Exploring Agentic Tool-Calling Decisions via Uncertainty-Aligned Reinforcement Learning2 days agoYijin Zhou, Linqian Zeng, Xiaoya Lu, Wenyuan Xie, Dongrui Liu, Junchi Yan, Jing Shao

Lean4Agent: Formal Modeling and Verification for Agent Workflow and Trajectory2 days agoRuida Wang, Jerry Huang, Pengcheng Wang, Xuanqing Liu, Luyang Kong, Tong Zhang

Contract2Tool: Learning Preconditions and Effects for Reliable Tool-Augmented LLM Agents2 days agoRahul Suresh Babu, Laxmipriya Ganesh Iyer

From Human Guidance to Autonomy: Agent Skill System for End-to-End LLM Deployment on Spatial NPUs2 days agoJiajie Li, Erwei Wang, Zhiru Zhang, Samuel Bayliss

Efficient Skill Grounding via Code Refactoring with Small Language Models2 days agoSera Choi, Wonje Choi, Saehun Chun, Daehee Lee, Jooyoung Kim, Chaeun Lee, Honguk Woo

Workflow-to-Skill: Skill Creation via Routing-Workflow-Semantics-Attachments Decomposition2 days agoYuyang Zhang, Xinyuan Han, Xudong Jiang, Run Wang

Declarative Skills for AI Agents in Knowledge-Grounded Tool-Use Workflows2 days agoM. Danish Lim, I. Danial Bin Sharudin, Wen Han Chen, Cedric Lim, Laura Wynter

Arxiv CS.AI (top papers)·Amine El HattamiNicolas ChapadosChristopher Pal

7 days ago