CommitLLM

CommitLLM is a verification framework designed to ensure the integrity and authenticity of large language model responses by generating compact receipts that record the model checkpoint, deployment configuration, decoding policies, and the specific execution path leading to the delivered answer, which can be independently verified on CPU. Its main features include binding input prompts, sampling randomness, output formatting, and decoding procedures into a tamper-resistant proof, enabling solid audit mechanisms through routine and deep verification modes, and supporting deployment alongside existing GPU serving stacks without kernel modifications, aimed at organizations requiring secure, auditable AI inference, such as enterprises, regulated industries, and decentralized marketplaces. It addresses problems of trust, model provenance, answer tampering, and compliance by providing cryptographically sound evidence that the response was produced by.

CommitLLM poate fi găsit în Contract Lifecycle Management (CLM) categorii.