Meta-Skill: Frozen-Weight Builder Models Learn Better Agent Harnesses
apodex · hf · 2026-10-01
This test-time AI-for-AI work has a frozen Builder model learn principles (Meta-Skills) for constructing better execution environments for a Target model, based on the Target's execution feedback. Full-bank meta-skills improve macro-average performance by 8.95 points over no-skill construction and 12.02 points over handing the same bank directly to the Target, suggesting a path to system-level self-improvement.
More from coding & agent
- Dev says ChatGPT plugin store blocks digital services and credit sales — NousResearch · 2026-10-01
- Dev Burns Through Claude Weekly Quota in Under Two Days on Opus 4.5 Alone — yihui_indie · 2026-10-01
- Split coding workflow: Opus 5.5 xhigh for planning, Sol 6.1 high for the bulk of coding — haider1 · 2026-10-01
- FV expert: AI folks' view of formal verification is 15 years out of date — tianyin_xu · 2026-10-01
- a16z's Eric Jang: Codex Computer Control Beats Email APIs for Managing Inboxes — ericjang11 · 2026-10-01
- Why Agentic Inference Needs P/D Disaggregation: Decoding 300K-Token Codebase Latency — abhijithneil · 2026-10-01