cd /news/machine-learning/self-improvement-memory-enables-cont… · home › topics › machine-learning › article
[ARTICLE · art-171352] src=arxiv.org ↗ pub= topic=machine-learning verified=true sentiment=↑ positive

Self-Improvement Memory Enables Continual Recursive Self-Improvement

A paper submitted to arXiv on 13 Nov 2026 proposes Self-Improvement Memory (SIM), a memory mechanism for the harness optimizer rather than the task-solving policy, to support Continual Recursive Self-Improvement (CRSI) across compatible benchmarks or policy models. SIM groups edits by recurring harness problem and records their rationale, predicted and observed outcomes, and evaluation context; across four domains it beat harness-only and root-cause-only continuation in cross-dataset and pairwise cross-model comparisons. In coding, across ten independent source runs, paired continuations with SIM raised final scores by an average of 2.36 points over harness-only continuation (p=0.003).

read2 min views7 publishedNov 17, 2026
Self-Improvement Memory Enables Continual Recursive Self-Improvement
Image: source
  [Submitted on 13 Nov 2026]


[View PDF](https://arxiv.org/pdf/2611.18856)

[HTML (experimental)](https://arxiv.org/html/2611.18856v1)

Abstract: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 extend this history as optimization continues. In our experiments across four domains, SIM achieves higher final scores than harness-only and root-cause-only continuation in cross-dataset and pairwise cross-model comparisons. Across ten independent source runs in coding, paired continuations with SIM improve final scores by an average of 2.36 points over harness-only continuation ($p=0.003$).

Additional Features

References & Citations

...

Bibliographic Explorer

(What is the Explorer?) Connected Papers

(What is Connected Papers?) Litmaps

(What is Litmaps?) scite Smart Citations

(What are Smart Citations?) alphaXiv

(What is alphaXiv?) CatalyzeX Code Finder for Papers

(What is CatalyzeX?) DagsHub

(What is DagsHub?) Gotit.pub

(What is GotitPub?) Hugging Face

(What is Huggingface?) ScienceCast

(What is ScienceCast?) Influence Flower

(What are Influence Flowers?) CORE Recommender

(What is CORE?) IArxiv Recommender

(What is IArxiv?) arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

── more in #machine-learning 4 stories · sorted by recency
── more on @arxiv 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
→ Live at https://your-agent.zahid.host ✓
Get free account → Pricing
from €0/mo · no card required
LIVE [news/self-improvement-mem…] indexed:0 read:2min 2026-11-17 · —