{"slug": "coremem-cost-internalizing-memory-for-successive-coding-tasks-with-prospective", "title": "CoReMem: Cost-Internalizing Memory for Successive Coding Tasks with Prospective Valuation", "summary": "A paper submitted to arXiv on 12 November 2026 introduces CoReMem, a cost-internalizing memory framework for coding agents that raises a 27B actor's resolve rate on the ChainSWE benchmark from 21.0% to 25.7%. CoReMem uses prospective segment valuation to compare runs of the next task with and without individual trajectory segments from the same inherited repository state, building a compact memory around anchors such as repository locations and executable checks, and adds provenance-gated value credit during actor adaptation. The method also cuts steps by 17% and total cache-aware tokens by 11%, including memory-module calls, relative to the untrained actor model without memory.", "body_md": "# Computer Science > Artificial Intelligence\n\n  [Submitted on 12 Nov 2026]\n\n# Title:CoReMem: Cost-Internalizing Memory for Successive Coding Tasks with Prospective Valuation\n\n[View PDF](http://arxiv.org/pdf/2611.17316v1)\n\n[HTML (experimental)](https://arxiv.org/html/2611.17316v1)\n\nAbstract:Coding agents tackle successive tasks on an evolving repository, carrying forward edits and writing memory before the next task is known. Selecting memory for unknown next tasks requires balancing future usefulness against the costs of writing and reading it. The actor must also decide when acting on available memory saves effort as the task unfolds. We introduce Coding-Repository Memory (CoReMem), a cost-internalizing framework that guides memory design and actor adaptation through measured memory value on the next task. Prospective segment valuation compares runs of the next task with and without individual trajectory segments from the same inherited repository state, guiding a compact memory built around anchors: repository locations and executable checks. Actor adaptation estimates on-policy memory value within rollout groups with and without memory, using a reward that prioritizes resolution and penalizes step count in successful runs. Provenance-gated value credit assigns the additional value term to steps identified as using memory-supplied anchors, while every step retains outcome credit. On ChainSWE, CoReMem raises the 27B actor's resolve rate from 21.0% to 25.7%, with 17% fewer steps and 11% fewer total cache-aware tokens, including memory-module calls, relative to the untrained actor model without memory.\n    \n\n### Additional Features\n\n### Current browse context:\n\ncs.AI\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/coremem-cost-internalizing-memory-for-successive-coding-tasks-with-prospective", "canonical_source": "http://arxiv.org/abs/2611.17316v1", "published_at": "2026-11-13 15:44:32+00:00", "updated_at": "2026-11-13 16:22:12.555315+00:00", "lang": "en", "topics": ["ai-agents", "ai-research", "large-language-models", "ai-tools"], "entities": ["CoReMem", "ChainSWE", "arXiv"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/coremem-cost-internalizing-memory-for-successive-coding-tasks-with-prospective", "markdown": "https://wpnews.pro/news/coremem-cost-internalizing-memory-for-successive-coding-tasks-with-prospective.md", "text": "https://wpnews.pro/news/coremem-cost-internalizing-memory-for-successive-coding-tasks-with-prospective.txt", "jsonld": "https://wpnews.pro/news/coremem-cost-internalizing-memory-for-successive-coding-tasks-with-prospective.jsonld"}}