Continual recursive self-improvement lets harness optimizers reuse and revise prior edits across changing benchmarks or policy models by storing rationales, predicted and observed outcomes, and evaluation contexts in a self-improvement memory, improving final scores over continuation baselines.
Harness optimization is often studied with a fixed policy model and benchmark. When either changes, the optimizer must decide which parts of an evolved harness to reuse or revise, but the harness alone may not explain why earlier edits were made or how they performed. In this work, we study Continual Recursive Self-Improvement (CRSI), a setting for continuing existing harness-optimization methods across compatible benchmarks or policy models. We propose Self-Improvement Memory (SIM) for the harness optimizer, rather than the task-solving policy. SIM groups edits by recurring harness problem and records their rationale, predicted and observed outcomes, and evaluation context. New attempts…
Continual Recursive Self-Improvement (CRSI)Self-Improvement Memory (SIM)harness optimizerharness optimizationpolicy modeltask-solving policyharness-only continuationroot-cause-only continuationcross-dataset comparisonspairwise cross-model comparisons
- Upvotes
- 10
- Authors
- 10
- Published
- Nov 13, 2026
Summary and abstract excerpt via Hugging Face Papers ↗ · arXiv.