← 研踪 PaperTrailarXiv:2610.07476 · v2
Can Power Draw Constrain Covert Compute? Limits of Analogue Verification for AI Governance
首次提交 2026-10-06 06:40
版本更新 2026-11-17 14:14
雷达观测 2026-11-18 19:43
摘要
Frontier AI treaties or agreements on limiting computation require external verification; an external auditor must be able to confirm how much computation actually ran and that parties are adhering to the agreement. Analogue, off-chip measurements such as power draw provide an information channel for verification. It is unknown how well these analogue channels can constrain the amount of computation against an adversary who actively tries to subvert the audit. We derive a closed form for , the largest hidden computation a power trace cannot exclude, as a fraction of the declared machine capacity. Measurements on NVIDIA A100 GPUs give in the worst case, while an adversarial matched-energy schedule hides at least . Analogue power measurements alone therefore constrain compute weakly. Additional restrictions granted by the threat model, such as the ability of the verifier to re-execute the declared work at an observed operating point, let the verifier push down to in the maximally restricted case. This gives a quantitative estimate of what analogue measurements can contribute to compute verification.
作者备注:14 pages, 8 figures. v2: clarified the definition of $β$ and the relaxed bound; added related work on zero-knowledge inference proofs and proof-of-useful-work; results unchanged
观测到的版本
相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。
v2 · 2026-11-17 14:14 · 最新观测版本
Can Power Draw Constrain Covert Compute? Limits of Analogue Verification for AI Governance
Frontier AI treaties or agreements on limiting computation require external verification; an external auditor must be able to confirm how much computation actually ran and that parties are adhering to the agreement. Analogue, off-chip measurements such as power draw provide an information channel for verification. It is unknown how well these analogue channels can constrain the amount of computation against an adversary who actively tries to subvert the audit. We derive a closed form for , the largest hidden computation a power trace cannot exclude, as a fraction of the declared machine capacity. Measurements on NVIDIA A100 GPUs give in the worst case, while an adversarial matched-energy schedule hides at least . Analogue power measurements alone therefore constrain compute weakly. Additional restrictions granted by the threat model, such as the ability of the verifier to re-execute the declared work at an observed operating point, let the verifier push down to in the maximally restricted case. This gives a quantitative estimate of what analogue measurements can contribute to compute verification.
查看此版本解读 →查看此版本原文 ↗