TOPLOC: A Locality Sensitive Hashing Scheme for Trustless Verifiable Inference

Researchers proposed TOPLOC, a locality-sensitive hashing scheme for verifiable inference that detects unauthorized modifications to large language models with 100% accuracy and no false positives or negatives in empirical evaluations. The method reduces proof memory overhead by 1000x, requiring only 258 bytes per 32 tokens compared to 262 KB for Llama 3.1-8B-Instruct, enabling faster validation than original inference.

Computer Science Cryptography and Security Submitted on 27 Jan 2025 v1 https://arxiv.org/abs/2501.16007v1 , last revised 30 May 2025 this version, v2 Title:TOPLOC: A Locality Sensitive Hashing Scheme for Trustless Verifiable Inference View PDF https://arxiv.org/pdf/2501.16007 HTML experimental https://arxiv.org/html/2501.16007v2 Abstract:Large language models LLMs have proven to be very capable, but access to frontier models currently relies on inference providers. This introduces trust challenges: how can we be sure that the provider is using the model configuration they claim? We propose TOPLOC, a novel method for verifiable inference that addresses this problem. TOPLOC leverages a compact locality-sensitive hashing mechanism for intermediate activations, which can detect unauthorized modifications to models, prompts, or precision with 100% accuracy, achieving no false positives or negatives in our empirical evaluations. Our approach is robust across diverse hardware configurations, GPU types, and algebraic reorderings, which allows for validation speeds significantly faster than the original inference. By introducing a polynomial encoding scheme, TOPLOC minimizes the memory overhead of the generated proofs by $1000\times$, requiring only 258 bytes of storage per 32 new tokens, compared to the 262 KB requirement of storing the token embeddings directly for Llama 3.1-8B-Instruct. Our method empowers users to verify LLM inference computations efficiently, fostering greater trust and transparency in open ecosystems and laying a foundation for decentralized, verifiable and trustless AI services. Submission history From: Jack Min Ong Mr view email https://arxiv.org/show-email/c0684d24/2501.16007 Mon, 27 Jan 2025 12:46:45 UTC 478 KB \ v1\ https://arxiv.org/abs/2501.16007v1 v2 Fri, 30 May 2025 23:07:40 UTC 139 KB References & Citations Loading... Bibliographic and Citation Tools Bibliographic Explorer What is the Explorer? https://info.arxiv.org/labs/showcase.html arxiv-bibliographic-explorer Connected Papers What is Connected Papers? https://www.connectedpapers.com/about Litmaps What is Litmaps? https://www.litmaps.co/ scite Smart Citations What are Smart Citations? https://www.scite.ai/ Code, Data and Media Associated with this Article alphaXiv What is alphaXiv? https://alphaxiv.org/ CatalyzeX Code Finder for Papers What is CatalyzeX? https://www.catalyzex.com DagsHub What is DagsHub? https://dagshub.com/ Gotit.pub What is GotitPub? http://gotit.pub/faq Hugging Face What is Huggingface? https://huggingface.co/huggingface ScienceCast What is ScienceCast? https://sciencecast.org/welcome Demos Recommenders and Search Tools Influence Flower What are Influence Flowers? https://influencemap.cmlab.dev/ CORE Recommender What is CORE? https://core.ac.uk/services/recommender arXivLabs: experimental projects with community collaborators arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both 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. Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs https://info.arxiv.org/labs/index.html .