The Verification Tax: Fundamental Limits of AI Auditing in the Rare-Error Regime
Jason Z Wang · 2026
Read The Verification Tax: Fundamental Limits of AI Auditing in the Rare-Error Regime by Jason Z Wang (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-12951.txt.
Description
Jason Z Wang proves calibration checks hit a statistical floor. The famous 0.012 post-scaling error on CIFAR-100 sits below noise. Minimax rate scales as Lipschitz times error over samples, all to the one third power. Self-evaluation without labels gives exactly zero calibration information. Miscalibration turns undetectable below a sharp phase transition. Active querying removes the Lipschitz constant, reducing estimation to detection. Verification cost grows exponentially with pipeline depth. Tests across five benchmarks and six models confirm the limits.
Catalog record
| CREATOR | Jason Z Wang |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 69853 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.12951