{"slug": "error-certificates-for-kv-cache-eviction-via-randomized-design", "title": "Error Certificates for KV-Cache Eviction via Randomized Design", "summary": "A new study proves that deterministic KV-cache eviction cannot estimate attention-output error, while randomized eviction with a Poisson-sampled tail and Hájek correction yields a per-step error certificate with 0.97 empirical coverage at no accuracy cost. The researchers pre-registered seven claims, losing three, but found that the certificate separates cache-induced from inherent failures (AUC 0.73–0.75) and schedules recomputation better than random or confidence gating.", "body_md": "# Computer Science > Machine Learning\n\n[Submitted on 23 Jul 2026]\n\n# Title:Error Certificates for KV-Cache Eviction via Randomized Design\n\n[View PDF](/pdf/2607.21475)\n\n[HTML (experimental)](https://arxiv.org/html/2607.21475v1)\n\nAbstract:Deterministic KV-cache eviction keeps the top-$k$ tokens under an importance score and deletes the rest. We prove that this design cannot know what it destroyed: evicted values can be altered so that everything the serving system retains is unchanged while the true attention-output error grows arbitrarily, so no serving-time estimator of that error is consistent. Randomized eviction restores identifiability. With a Poisson-sampled tail at known inclusion probabilities, one logit offset performs the Hájek correction inside the softmax, and a survey-sampling variance estimator over the retained set becomes a per-step error certificate with 0.97 empirical coverage at no accuracy cost. On real workloads we pre-registered seven claims and lost three: question-aware eviction at 25--50\\% budgets is nearly free; output log-probability predicts failure better than the certificate; certificate-gated budget escalation adds nothing. What survives is attribution: the certificate separates cache-induced from inherent failures (AUC 0.73--0.75, against 0.47--0.54 for output confidence) and schedules recomputation better than random or confidence gating. Randomization buys attribution, not prediction.\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/))# 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))# 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))# 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/error-certificates-for-kv-cache-eviction-via-randomized-design", "canonical_source": "https://arxiv.org/abs/2607.21475", "published_at": "2026-07-26 06:07:06+00:00", "updated_at": "2026-07-26 06:22:40.989401+00:00", "lang": "en", "topics": ["machine-learning", "large-language-models", "ai-research"], "entities": [], "alternates": {"html": "https://wpnews.pro/news/error-certificates-for-kv-cache-eviction-via-randomized-design", "markdown": "https://wpnews.pro/news/error-certificates-for-kv-cache-eviction-via-randomized-design.md", "text": "https://wpnews.pro/news/error-certificates-for-kv-cache-eviction-via-randomized-design.txt", "jsonld": "https://wpnews.pro/news/error-certificates-for-kv-cache-eviction-via-randomized-design.jsonld"}}