{"version":1,"type":"story","url":"https://digestai.news/story/spem-framework-improves-skill-use-in-llm-agents","json":"https://digestai.news/story/spem-framework-improves-skill-use-in-llm-agents.json","markdown":"https://digestai.news/story/spem-framework-improves-skill-use-in-llm-agents.md","slug":"spem-framework-improves-skill-use-in-llm-agents","headline":"SPEM framework improves skill use in LLM agents","summary":"SPEM is a lightweight structure‑aware procedural memory framework designed to make skill reuse in large‑language‑model agents more efficient. The system stores individual skills in an Atomic Skill Memory and builds reusable skill compositions in a Skill‑DAG Memory. When a new task arrives, SPEM first looks for high‑confidence Skill‑DAG templates; if none are found, it retrieves atomic skills and uses a Skill‑GNN to select the appropriate ones based on state and structure. Successful runs are written back as new Skill‑DAGs, allowing the memory to grow over time.\n\nExperiments on the ALFWorld and SkillsBench benchmarks show that SPEM achieves the best overall effectiveness‑efficiency trade‑off, reaching 87.1 % success on ALFWorld and 38.38 % reward on SkillsBench while cutting token and runtime costs compared to strong baselines.","keyPoints":["SPEM uses Atomic Skill Memory and Skill‑DAG Memory to reduce token and runtime costs","SPEM reaches 87.1 % success on ALFWorld and 38.38 % reward on SkillsBench","It reuses high‑confidence Skill‑DAG templates and writes back new DAGs after successful runs"],"whyItMatters":"By making skill reuse cheaper and faster, SPEM could enable LLM agents to handle more complex tasks with fewer tokens, lowering costs and improving real‑world applicability.","category":{"slug":"agents","name":"Agents & Tools","url":"https://digestai.news/category/agents"},"entities":{"companies":[],"models":["SPEM"],"people":[]},"firstPublishedAt":"2026-10-21T04:00:00Z","updatedAt":"2026-10-21T04:00:00Z","sourceCount":1,"hasPrimarySource":true,"sources":[{"outlet":"arXiv cs.AI","title":"SPEM: Structure-Aware Procedural Memory for Efficient Skill Use in LLM Agents","url":"https://arxiv.org/abs/2610.26995","publishedAt":"2026-10-21T04:00:00Z","type":"primary","primary":true,"lead":true}],"sourceNotes":null,"discussions":[],"thread":null,"cite":{"text":"Digest AI, \"SPEM framework improves skill use in LLM agents\", 21 October 2026, https://digestai.news/story/spem-framework-improves-skill-use-in-llm-agents","publisher":"Digest AI","title":"SPEM framework improves skill use in LLM agents","datePublished":"2026-10-21T04:00:00Z","url":"https://digestai.news/story/spem-framework-improves-skill-use-in-llm-agents"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}