← 首页|学术|HyperSkill: Self-Evolving LLM Agents via Hypergraph-Structured Skill Memory
cs.CL · 2608.16114 · 2026/08/17

HyperSkill: Self-Evolving LLM Agents via Hypergraph-Structured Skill Memory

Xu, Ruiyao, Yang, Tiankai, Huang, Wei-Chieh
TL;DR:HyperSkill:超图结构技能记忆——subtask 步骤与可复用技能两类节点,每条轨迹的超边连接其涉及的 subtask 和技能;双路径检索(subtask 层+轨迹层)按共现排序技能,周期性结构知情维护剪枝低效用节点、按质量加权合并冗余技能。

🎯 问题

记忆设计要同时回答存什么、怎么组织检索、怎么演化;现有系统各只解决一部分——轨迹/洞察/工作流当孤立条目存、平面 embedding 相似度检索丢关系信号、维护不看结构。

🔬 方法

记忆 = 超图:两类节点(subtask steps, reusable skills)+ 每轨迹一条超边;双路径检索(subtask 与轨迹层)按检索轨迹上的共现给技能排序;周期性结构知情维护:剪低效用节点、按质量加权传播合并冗余技能。
章节结构(全文标题提取):
1 Introduction
2 Related Work
3 HyperSkill
4 Experiments
5 Conclusion
Limitations

📊 结果

在需要跨任务复用过程性知识的 agent 任务上,比平面轨迹存储与纯 embedding 检索的基线提升任务性能与记忆演化质量。

💡 与研究方向关联

技能记忆的「关系结构化」路线:把过程性知识从孤立轨迹里解放出来、显式建模组合关系——与 user 关注的 agent 技能自演化方向(配合 JailbreakSkill 等技能复用研究)直接相关。

📝 原文摘要

▶ 原文摘要 Abstract
As agentic tasks grow in complexity, LLM agents increasingly rely on experiential memory to reuse procedural knowledge across tasks. Effective memory design must jointly address what to store, how memory is structured and retrieved, and how memory evolves. Existing systems tackle each only partially: they store trajectories, insights, or workflows as isolated entries, discarding compositional relationships among subtasks and reusable skills; retrieve by flat embedding similarity that ignores relational signals; and maintain memory without leveraging its relational structure. We propose HyperSkill, a hypergraph-based memory framework that jointly improves all three. HyperSkill represents memory as a hypergraph with two node types, subtask steps and reusable skills, where each hyperedge links the subtasks and skills from a single trajectory. Dual-path retrieval queries both subtask and trajectory levels, ranking skills by co-occurrence across retrieved trajectories. Periodic structure-informed maintenance prunes low-utility nodes and merges redundant skills via quality-weighted propagation. Across xBench, GAIA, and WebWalkerQA with GPT-4o and Qwen3-30B-A3B, HyperSkill outperforms ten memory baselines, yielding gains of up to +11.51 on GAIA and +11.18 on WebWalkerQA.
Deep Read · 2026-08-19高松灯 / Agent 日报 · 会话与记忆
技能记忆超图自演化检索