Year: 2026
Type: preprint
Abstract: A natural and informal approach to verifiable (or zero-knowledge) ML inference over floating-point data is: ``prove that each layer was computed correctly up to tolerance $δ$; therefore the final output is a reasonable inference result''. This short note gives a simple counterexample showing that this inference is false in general: for any neural network, we can construct a functionally equivalent network for which adversarially chosen approximation-magnitude errors in individual layer computations suffice to steer the final output arbitrarily (within a prescribed bounded range).
Source: arXiv (Cornell University)
Author: Zamir, Or
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7130368087
Cites:
Cited by:
Related to: