{"slug": "kvguard-a-kv-cache-safety-checker-for-llm-serving-systems", "title": "KVGuard: A KV Cache Safety Checker for LLM Serving Systems", "summary": "A November 4, 2026 arXiv paper presents KVGuard, a memory-safety checker for KV cache management in LLM serving systems that catches double-free, use-after-free, out-of-bounds access, uninitialized read and permission violations which existing CUDA-level sanitizers miss. A prototype on vLLM and SGLang covered all 29 historical real-world KV cache safety bugs from the two systems' patch histories and found 8 previously unknown violations with no false positives, adding 3.0% and 7.4% geomean TPOT overhead on Qwen 2.5 14B and at most 0.11% KV cache memory overhead — two orders of magnitude below NVIDIA Compute Sanitizer on the same workloads. KVGuard tracks a compact shadow block per physical KV block recording per-request permission and per-entry initialization, firing each check at the boundary that introduced the violation rather than reporting downstream output corruption many requests later.", "body_md": "# Computer Science > Cryptography and Security\n\n  [Submitted on 4 Nov 2026]\n\n# Title:KVGuard: A KV Cache Safety Checker for LLM Serving Systems\n\n[View PDF](https://arxiv.org/pdf/2611.06138)\n\n[HTML (experimental)](https://arxiv.org/html/2611.06138v1)\n\nAbstract:Modern LLM serving systems manage the KV cache through PagedAttention, a custom GPU allocator that partitions a single CUDA-allocated region into fixed-size physical blocks and maps each request to those blocks through a CPU-built block table. This block-based design exposes a previously unstudied class of memory safety bugs in the KV cache, including double-free, use-after-free, out-of-bounds access, uninitialized read, and permission violation, which silently corrupt outputs or break request isolation without ever crashing the system. Existing memory sanitizers cannot detect them, because they reason about CUDA-level addresses rather than KV cache blocks, requests, and block tables, and their redzone- and quarantine-based designs do not fit the scarce GPU memory budget of LLM serving. This paper presents KVGuard, the first systematic memory-safety abstraction for KV cache management in LLM serving systems. KVGuard maintains a compact shadow block for each physical KV block, recording per-request permission and per-entry initialization, and uses this metadata to perform five safety checks across the PagedAttention workflow, spanning CPU allocation, block table construction and translation, and GPU-side block table lookup and KV entry access. Each check fires at the boundary that introduced the violation, enabling early detection during development rather than reporting downstream output corruption many requests later. A prototype of KVGuard on vLLM and SGLang covers all 29 historical real-world KV cache safety bugs collected from the two systems' patch histories. KVGuard further found 8 previously unknown safety violations while reporting no false positives. It adds 3.0% and 7.4% geomean TPOT overhead on Qwen 2.5 14B on vLLM and SGLang and at most 0.11% of the KV cache in memory overhead, two orders of magnitude below NVIDIA Compute Sanitizer on the same workloads.\n    \n\n### Additional Features\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/kvguard-a-kv-cache-safety-checker-for-llm-serving-systems", "canonical_source": "https://arxiv.org/abs/2611.06138", "published_at": "2026-11-05 10:22:21+00:00", "updated_at": "2026-11-05 10:56:02.156779+00:00", "lang": "en", "topics": ["ai-safety", "ai-infrastructure", "large-language-models", "ai-research", "mlops"], "entities": ["KVGuard", "vLLM", "SGLang", "Qwen 2.5 14B", "NVIDIA Compute Sanitizer", "PagedAttention", "arXiv"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/kvguard-a-kv-cache-safety-checker-for-llm-serving-systems", "markdown": "https://wpnews.pro/news/kvguard-a-kv-cache-safety-checker-for-llm-serving-systems.md", "text": "https://wpnews.pro/news/kvguard-a-kv-cache-safety-checker-for-llm-serving-systems.txt", "jsonld": "https://wpnews.pro/news/kvguard-a-kv-cache-safety-checker-for-llm-serving-systems.jsonld"}}