AIDE²: Self-Improving Research Agent Beats Two Years of Human Tuning

The AIDE² paper from WecoAI describes a recursive self-improvement system where an AI research agent modifies its own code, iterating seven times in eight days and ultimately outperforming an agent hand-tuned over two years.

2026-09-23 ~ 2026-09-24 · 3 related posts

• AIDE^2 shows AI research agents can recursively improve themselves over 8 days — WecoAI · 2026-09-23
• AIDE² paper: AI research agent self-improves for 8 days, beats 2-year hand-tuned harness — ptkbhv · 2026-09-24

1 near-duplicate retellings: egrefen