{"version":1,"type":"story","url":"https://digestai.news/story/self-direction-lets-agents-learn-from-mistakes-boosting-success-rates","json":"https://digestai.news/story/self-direction-lets-agents-learn-from-mistakes-boosting-success-rates.json","markdown":"https://digestai.news/story/self-direction-lets-agents-learn-from-mistakes-boosting-success-rates.md","slug":"self-direction-lets-agents-learn-from-mistakes-boosting-success-rates","headline":"Self-Direction lets agents learn from mistakes, boosting success rates","summary":"Self‑Direction is a new mechanism that lets agents learn from their own mistakes without expensive fine‑tuning. When an agent fails, it reflects on the trajectory, extracts a lesson, and distills it into a lightweight low‑rank adapter. The adapter adds only about 0.02% of the base model’s parameters and keeps the original weights frozen.\n\nThe authors tested Self‑Direction on Qwen family models across three benchmarks: GAIA for open‑ended research, Claw‑Eval for tool‑use workflows, and Terminal‑Bench 2.1 for terminal tasks. The method raised overall success rates by 8.9 to 17.0 percentage points and nearly doubled the reliability of an agent’s corrections. Learned adapters were stored in a persistent library and could generalise to dozens of new tasks without further training.\n\nThis approach offers a persistent, low‑cost way for agents to remember lessons across sessions, reducing the need for full‑parameter fine‑tuning and making open‑weight models more reliable in real‑world applications.","keyPoints":["Self‑Direction uses low‑rank adapters that add only ~0.02% of base model parameters.","Experiments show 8.9–17.0 percentage‑point improvement in success rates.","Adapters generalise to dozens of new tasks without extra training."],"whyItMatters":"By enabling agents to retain lessons across sessions, Self‑Direction cuts the cost of fine‑tuning and improves reliability, accelerating the deployment of open‑weight AI agents.","category":{"slug":"research","name":"Research","url":"https://digestai.news/category/research"},"entities":{"companies":[],"models":["Qwen"],"people":[]},"firstPublishedAt":"2026-10-26T04:00:00Z","updatedAt":"2026-10-26T04:00:00Z","sourceCount":2,"hasPrimarySource":true,"sources":[{"outlet":"arXiv cs.AI","title":"Don't Forget What You Just Learned: Self-Directed Repair Turns Agent Mistakes Into Lasting Lessons","url":"https://arxiv.org/abs/2610.31846","publishedAt":"2026-10-26T04:00:00Z","type":"primary","primary":true,"lead":true},{"outlet":"arXiv cs.AI","title":"RSI-Opt: Recursive Self-Improvement Optimizer Agents For Model Training","url":"https://arxiv.org/abs/2610.32026","publishedAt":"2026-10-26T04:00:00Z","type":"primary","primary":true,"lead":false}],"sourceNotes":null,"discussions":[],"thread":{"title":"AI Audio Arms Race Accelerates with Low‑Cost Real‑Time Voice","url":"https://digestai.news/thread/nari-unveils-qwen3tts-and-qwen3asr-with-low-latency-and-competitive-pricing","storyCount":6},"cite":{"text":"Digest AI, \"Self-Direction lets agents learn from mistakes, boosting success rates\", 26 October 2026, https://digestai.news/story/self-direction-lets-agents-learn-from-mistakes-boosting-success-rates","publisher":"Digest AI","title":"Self-Direction lets agents learn from mistakes, boosting success rates","datePublished":"2026-10-26T04:00:00Z","url":"https://digestai.news/story/self-direction-lets-agents-learn-from-mistakes-boosting-success-rates"},"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"}