Recursive Scales Parallel Experiments to 100K with Steady Gains

Recursive SI argues that open-endedness is key to automated research, running massive parallel experiments like a community of agent scientists. On a kernel benchmark, scaling experiments from 1,000 to 100,000 yielded continuous improvements in research progress.

2026-11-12 ~ 2026-11-12 · 2 related posts

• Recursive bets on open-endedness: agents running massive parallel experiments like scientist communities — agihouse_org · 2026-11-12
• Kernel benchmark keeps improving as Recursive scales experiments from 1,000 to 100,000 — agihouse_org · 2026-11-12