SOURCE-LINKED INTELLIGENCE

RRSI: Regularized Recursive Self-Improvement of Agent Harnesses

arXiv · AI, language, vision and robotics · article · Sep 21, 2026 · UTC

An LLM agent's capability is largely magnified by its harness, namely the prompts, control flow, tooling, memory, and context management surrounding the frozen backbone model. Recent methods increasingly automate this process by iteratively proposing and selecting component-wise edits of an agent harness, practically establishing a form of recursive self-improvement (RSI) at the agent-system level. However, such recursive evolution may overfit by memorizing the training tasks, showing large in-distribution gains that shrink or even vanish on out-of-distribution benchmarks. We introduce Regular

Read original source ↗ Open in workspace

recordType

paper

region

Global