Anthropic open-sources FlashProve, its provable inference prototype for MoE models

November 4, 2026 · 4:36 am 1 sources

Image: Help Net Security

Coverage

Coverage

• Anthropic open-sources FlashProve, its provable inference prototype for MoE models
  Help Net Security · November 4, 2026 04:36
  Anthropic’s Frontier Red Team, with collaborators at EPFL, MIT and StarkWare, built FlashProve, a prototype that produces a cryptographic proof that one specific language model generated one specific reply. The prover, two verifiers, the formal proofs behind them and a protocol specification are…