{"slug": "dimwit-directly-invoking-models-without-intensive-tuning", "title": "DIMWIT: Directly Invoking Models Without Intensive Tuning", "summary": "A paper submitted to arXiv on 1 November 2026 proposes DIMWIT, an approach that skips task-specific policy development and instead asks pretrained models to make resource-management decisions directly from the current system state, desired outcome, and available choices. Evaluated on microservice autoscaling, DRAM sizing, and video encoding with a frontier LLM and two small decision models, a small decision model met latency targets using within 10% of the replicas allocated by policy-based autoscalers and within 13% of the frontier LLM, cut DRAM usage by up to 27.5% versus static provisioning while keeping at least 95% of baseline performance, and completed video encodes up to 4.3x faster than the frontier LLM with a 1.1-point VMAF drop. The authors conclude that making individual resource-management decisions may be considerably simpler than developing a general strategy for making them.", "body_md": "# Computer Science > Performance\n\n  [Submitted on 1 Nov 2026]\n\n# Title:DIMWIT: Directly Invoking Models Without Intensive Tuning\n\n[View PDF](https://arxiv.org/pdf/2611.01463)\n\n[HTML (experimental)](https://arxiv.org/html/2611.01463v1)\n\nAbstract:Resource managers typically encode what to do in a task-specific policy---a heuristic, an optimization procedure, or a model trained for the task---that must be developed before use and may need revision as the workload or system changes. We propose bypassing policy development and instead asking pretrained models to make resource-management decisions directly, given the current system state, desired outcome, and available choices. We evaluate this approach on microservice autoscaling, DRAM sizing, and video encoding, using a frontier LLM and two small decision models. On two microservice applications, a small decision model meets latency targets while using within 10% of the replicas allocated by policy-based autoscalers and within 13% of the frontier LLM. In DRAM sizing, a small decision model reduces DRAM usage by up to 27.5% relative to static provisioning while maintaining at least 95% of baseline performance. In video encoding, it completes encodes up to 4.3$\\times$ faster than the frontier LLM, with an imperceptible 1.1-point decrease in Video Multi-Method Assessment Fusion (VMAF), a metric of perceptual video quality. Making individual resource-management decisions may be considerably simpler than developing a general strategy for making them.\n    \n\n### Additional Features\n\n### Current browse context:\n\ncs.PF\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/dimwit-directly-invoking-models-without-intensive-tuning", "canonical_source": "https://arxiv.org/abs/2611.01463", "published_at": "2026-11-11 19:17:35+00:00", "updated_at": "2026-11-11 19:51:20.699377+00:00", "lang": "en", "topics": ["machine-learning", "ai-research", "ai-infrastructure", "large-language-models"], "entities": ["DIMWIT", "arXiv", "VMAF"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/dimwit-directly-invoking-models-without-intensive-tuning", "markdown": "https://wpnews.pro/news/dimwit-directly-invoking-models-without-intensive-tuning.md", "text": "https://wpnews.pro/news/dimwit-directly-invoking-models-without-intensive-tuning.txt", "jsonld": "https://wpnews.pro/news/dimwit-directly-invoking-models-without-intensive-tuning.jsonld"}}