RSIAgent Achieves Recursive Self-Improvement Without Training

AetherAI released RSIAgent, a training-free multi-agent framework that enables recursive self-improvement by scaling experience instead of model parameters, allowing open-source agents to surpass GPT-6 Astra on two agent benchmarks.

2026-09-15 ~ 2026-09-16 · 3 related posts

• RSIAgent: parameter-free self-improvement pushes open-source models past GPT-6 Astra on agent benchmarks — 机器之心 · 2026-09-15
• RSIAgent: training-free multi-agent framework for recursive self-improvement in new environments — AetherLabs-AI · 2026-09-15
• RSIAgent improves agents without weight updates, beats GPT-6 Astra on OSWorld 2.0 — Roger_M_Taylor · 2026-09-16