{"slug": "stepped-moe-reducing-memory-movement-in-large-mixture-of-experts-models", "title": "Stepped MoE: reducing memory movement in large mixture-of-experts models", "summary": "A paper submitted to arXiv on 5 Oct 2026 introduces \"Stepped MoE,\" a unified framework that combines elastic nested sub-networks with sparsely gated mixture-of-experts routing so a single model can serve at 1, 2, 3, or 4 billion parameters at inference time. The authors report the elastic model is 2-5% more accurate than dense counterparts on knowledge-intensive benchmarks, matches its static versions, and delivers latency similar to dense models while sharing parameters to save on-device disk space.", "body_md": "# Computer Science > Machine Learning\n\n  [Submitted on 5 Oct 2026]\n\n# Title:Stepped MoE: Segment-Level Routing with Configurable Inference Complexity\n\n[View PDF](https://arxiv.org/pdf/2610.07348)\n\n[HTML (experimental)](https://arxiv.org/html/2610.07348v1)\n\nAbstract:Training large language models (LLMs) is resource-intensive, and adapting them for diverse deployment scenarios with varying computational constraints remains challenging. While elastic architectures enable flexible model deployment and sparsely activated models allow input-adaptive computation, existing approaches treat these dimensions independently. Moreover, models catered towards on-device edge inference need to conform to the memory and compute limitations of the serving devices. In this paper, we introduce a unified framework that combines elastic structures with sparsely gated architectures to create models that adapt simultaneously to both deployment constraints and task requirements. Our approach employs a model backbone that conditions on both the context and target efficiency specifications, enabling fine-grained control over the accuracy-efficiency trade-off at inference time. The model learns to activate task-relevant parameters within elastically-nested sub-networks, allowing a single model to span multiple capacity points while maintaining input-adaptive routing. Through experiments we demonstrate that we can create a model that allows the flexibility to use 1,2,3,4 billion parameters while being more accurate than their dense counter-parts (2-5\\% on knowledge-intensive benchmarks) and at par with their static versions while delivering similar latency metrics as dense models. Overall, we save on device disk space by sharing the model parameters, allow flexibility of serving based on DRAM and compute available while delivering more accurate results.\n    \n\n### Additional Features\n\n### Current browse context:\n\ncs.LG\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))\nIArxiv Recommender\n\n*(*[What is IArxiv?](https://iarxiv.org/about))\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/stepped-moe-reducing-memory-movement-in-large-mixture-of-experts-models", "canonical_source": "https://arxiv.org/abs/2610.07348", "published_at": "2026-10-07 16:11:35+00:00", "updated_at": "2026-10-07 16:20:46.371548+00:00", "lang": "en", "topics": ["machine-learning", "large-language-models", "ai-research", "ai-infrastructure"], "entities": ["arXiv", "Stepped MoE"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/stepped-moe-reducing-memory-movement-in-large-mixture-of-experts-models", "markdown": "https://wpnews.pro/news/stepped-moe-reducing-memory-movement-in-large-mixture-of-experts-models.md", "text": "https://wpnews.pro/news/stepped-moe-reducing-memory-movement-in-large-mixture-of-experts-models.txt", "jsonld": "https://wpnews.pro/news/stepped-moe-reducing-memory-movement-in-large-mixture-of-experts-models.jsonld"}}