{"slug": "harness-learning-enables-generalizable-test-time-adaptation", "title": "Harness Learning Enables Generalizable Test-Time Adaptation", "summary": "A paper submitted to arXiv on 28 Sep 2026 introduces \"harness learning,\" a method that trains a proposer model via reinforcement learning to revise a solver language-model agent's harness — the executable program organizing model calls, tool use, and information flow — using execution feedback as the reward. The authors formulate the process as meta-learning over executable programs, with harness revisions playing the role of weight updates in gradient-based adaptation, and report that at test time the proposer refines the harness from successive executions on a new task without any parameter-space update. Experiments on reasoning and multi-hop question answering show harness learning improves revision quality and that test-time adaptation ability transfers to unseen tasks, with policies trained on individual revisions continuing to improve harnesses over multiple rounds.", "body_md": "# Computer Science > Computation and Language\n\n  [Submitted on 28 Sep 2026]\n\n# Title:Harness Learning Enables Generalizable Test-Time Adaptation\n\n[View PDF](https://arxiv.org/pdf/2609.35738)\n\n[HTML (experimental)](https://arxiv.org/html/2609.35738v1)\n\nAbstract:A language-model agent is jointly defined by its model and its harness, the executable program that organizes model calls, tool use, and information flow. Because different tasks call for different ways of organizing these operations, the harness needs to be adapted using feedback from the task at hand. We introduce harness learning, which trains a proposer model to revise a solver's harness using execution feedback. We formulate this process as meta-learning over executable programs, with harness revisions playing the role of weight updates in gradient-based adaptation. We train the proposer with reinforcement learning, using the task performance of revised harnesses as the reward. At test time, the proposer uses feedback from successive executions on a new task to refine the harness, without performing any parameter-space update. Experiments on reasoning and multi-hop question answering show that harness learning improves revision quality and that the ability to adapt at test time transfers to unseen tasks. Policies trained on individual revisions can continue improving harnesses over multiple rounds, while the benefits of training on revision sequences vary across settings. These findings suggest a path towards continually learning agents that turn accumulated experience into generalizable improvements.\n    \n\n### References & Citations\n\nLoading...\n\n# Bibliographic and Citation Tools\n\nBibliographic Explorer \n\n*(*[What is the Explorer?](https://info.arxiv.org/labs/showcase.html#arxiv-bibliographic-explorer))\nConnected Papers \n\n*(*[What is Connected Papers?](https://www.connectedpapers.com/about))\nLitmaps \n\n*(*[What is Litmaps?](https://www.litmaps.co/))\nscite Smart Citations \n\n*(*[What are Smart Citations?](https://www.scite.ai/))\n# Code, Data and Media Associated with this Article\n\nalphaXiv \n\n*(*[What is alphaXiv?](https://alphaxiv.org/))\nCatalyzeX Code Finder for Papers \n\n*(*[What is CatalyzeX?](https://www.catalyzex.com))\nDagsHub \n\n*(*[What is DagsHub?](https://dagshub.com/))\nGotit.pub \n\n*(*[What is GotitPub?](http://gotit.pub/faq))\nHugging Face \n\n*(*[What is Huggingface?](https://huggingface.co/huggingface))\nScienceCast \n\n*(*[What is ScienceCast?](https://sciencecast.org/welcome))\n# Demos\n\n# Recommenders and Search Tools\n\nInfluence Flower \n\n*(*[What are Influence Flowers?](https://influencemap.cmlab.dev/))\nCORE Recommender \n\n*(*[What is CORE?](https://core.ac.uk/services/recommender))\n# arXivLabs: experimental projects with community collaborators\n\narXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.\n\nBoth 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.\n\nHave an idea for a project that will add value for arXiv's community? [**Learn more about arXivLabs**](https://info.arxiv.org/labs/index.html).", "url": "https://wpnews.pro/news/harness-learning-enables-generalizable-test-time-adaptation", "canonical_source": "https://arxiv.org/abs/2609.35738", "published_at": "2026-10-12 05:31:58+00:00", "updated_at": "2026-10-12 05:38:54.594178+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-agents", "ai-research", "machine-learning"], "entities": ["arXiv", "harness learning"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/harness-learning-enables-generalizable-test-time-adaptation", "markdown": "https://wpnews.pro/news/harness-learning-enables-generalizable-test-time-adaptation.md", "text": "https://wpnews.pro/news/harness-learning-enables-generalizable-test-time-adaptation.txt", "jsonld": "https://wpnews.pro/news/harness-learning-enables-generalizable-test-time-adaptation.jsonld"}}