anthropic.com2 days agoProgress on provable inference for large language modelsSharing progress on a powerful type of cryptographic proof known as a zkSNARK, which is uniquely fit for provable inference.Visit anthropic.comBookmarkAdd to collection