ModularRSI: Modular, benchmark-disjoint framework for generalizable agent harness self-improvement

IQuestLab · hf · 2026-09-16

IQuestLab introduces ModularRSI, a framework for generalizable recursive self-improvement (RSI) on agent harnesses, tackling three known failures: benchmark-specific adaptation, noisy single-trajectory updates, and hard-to-attribute monolithic optimization.

Key ideas:

Experiments on TB2.0 and SWE-Bench Verified show consistent improvements on unseen in-domain and cross-domain tasks, with the evolved harness transferring across different foundation models.

Original post →

More from coding & agent

coding & agent channel →