Recurrent Harness for Agentic Continual Learning

Hugging Face Daily Papers Papers
continual-learning agents memory-revision language-models harness online-adaptation

Summary

论文提出ReHarness，一种用于智能体持续学习的循环harness框架，在严格在线流式协议下通过backward revision补充前向信息流，从而在分布偏移下改进语言模型智能体的记忆与反思更新。

Language model agents increasingly operate over long interaction streams, requiring online adaptation to distribution shifts through continual learning. However, existing agent harnesses typically either evolve their states only through forward information flow or rely on additional replay or validation data for harness optimization. To address this, we propose ReHarness, a recurrent harness framework for agentic continual learning that complements forward read and write with backward revision under a strictly online streaming protocol. We ask what information could have been captured differently in the harness state to better inform the current instance in the stream. Specifically, after each instance, we leverage the resulting trajectory as feedback to revise the reflections retrieved for that instance, thereby propagating feedback backward through the textual harness state. Across three continual learning benchmarks and three language model backbones, ReHarness consistently outperforms strong raw-context baselines, memory- and reflection-based methods, and prompt optimization approaches. Further analysis shows that ReHarness produces source-grounded and meaningful revisions, corrects overgeneralized or outdated experience under distribution shifts, and improves both stability and plasticity during continual learning.
Original Article
View Cached Full Text

Cached at: 10/12/26, 06:54 AM

Paper page - Recurrent Harness for Agentic Continual Learning

Source: https://huggingface.co/papers/2610.12668

Abstract

Languagemodelagentsincreasinglyoperateoverlonginteractionstreams,requiringonlineadaptationtodistributionshiftsthroughcontinuallearning.However,existingagentharnessestypicallyeitherevolvetheirstatesonlythroughforwardinformationfloworrelyonadditionalreplayorvalidationdataforharnessoptimization.Toaddressthis,weproposeReHarness,arecurrentharnessframeworkforagenticcontinuallearningthatcomplementsforwardreadandwritewithbackwardrevisionunderastrictlyonlinestreamingprotocol.Weaskwhatinformationcouldhavebeencaptureddifferentlyintheharnessstatetobetterinformthecurrentinstanceinthestream.Specifically,aftereachinstance,weleveragetheresultingtrajectoryasfeedbacktorevisethereflectionsretrievedforthatinstance,therebypropagatingfeedbackbackwardthroughthetextualharnessstate.Acrossthreecontinuallearningbenchmarksandthreelanguagemodelbackbones,ReHarnessconsistentlyoutperformsstrongraw-contextbaselines,memory-andreflection-basedmethods,andpromptoptimizationapproaches.FurtheranalysisshowsthatReHarnessproducessource-groundedandmeaningfulrevisions,correctsovergeneralizedoroutdatedexperienceunderdistributionshifts,andimprovesbothstabilityandplasticityduringcontinuallearning.

View arXiv pageView PDFAdd to collection

Get this paper in your agent:

hf papers read 2610\.12668

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2610.12668 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2610.12668 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2610.12668 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

Recurrent Harness for Agentic Continual Learning

arXiv cs.LG

提出 ReHarness 框架，通过在严格在线流式协议下结合前向读写与后向修订（利用轨迹反馈修订检索到的反思），实现 LLM 智能体的持续学习，在三个基准和三种模型骨干上均优于基线方法。

Self-Evolving Agent Harness

Hugging Face Daily Papers

提出一个自我进化代理框架（Harness RSI），在保持模型权重固定的情况下，通过反馈驱动的修订与进化优化来持续改进代理的软件脚手架（提示、记忆、工作流和可执行组件），并借助持续 PAC 学习给出泛化与保留的理论保证。

Continual Harness: Online Adaptation for Self-Improving Foundation Agents

Hugging Face Daily Papers

The paper introduces 'Continual Harness,' a framework enabling embodied AI agents to self-improve online without environment resets. It demonstrates significant progress in playing Pokémon games, achieving human-level performance through automated prompt and skill refinement.

Google Harness² for Recursive Self Improvement (GitHub Repo)

TLDR AI

Google Research 开源 Harness²，一个实现递归式智能体系统优化（recursive agent harnessing）的即插即用框架，可在模型权重固定的情况下通过 Propose–Probe–Compose 循环持续改进 agent harness，在多个专业基准上带来最高 11.8 个百分点的性能提升。