ResearchTrend.AI
  • Papers
  • Communities
  • Events
  • Blog
  • Pricing
Papers
Communities
Social Events
Terms and Conditions
Pricing
Parameter LabParameter LabTwitterGitHubLinkedInBlueskyYoutube

© 2025 ResearchTrend.AI, All rights reserved.

  1. Home
  2. Papers
  3. 2306.04505
9
1

Hardness of Deceptive Certificate Selection

7 June 2023
Stephan Wäldchen
    AAML
ArXivPDFHTML
Abstract

Recent progress towards theoretical interpretability guarantees for AI has been made with classifiers that are based on interactive proof systems. A prover selects a certificate from the datapoint and sends it to a verifier who decides the class. In the context of machine learning, such a certificate can be a feature that is informative of the class. For a setup with high soundness and completeness, the exchanged certificates must have a high mutual information with the true class of the datapoint. However, this guarantee relies on a bound on the Asymmetric Feature Correlation of the dataset, a property that so far is difficult to estimate for high-dimensional data. It was conjectured in W\"aldchen et al. that it is computationally hard to exploit the AFC, which is what we prove here. We consider a malicious prover-verifier duo that aims to exploit the AFC to achieve high completeness and soundness while using uninformative certificates. We show that this task is NP\mathsf{NP}NP-hard and cannot be approximated better than O(m1/8−ϵ)\mathcal{O}(m^{1/8 - \epsilon})O(m1/8−ϵ), where mmm is the number of possible certificates, for ϵ>0\epsilon>0ϵ>0 under the Dense-vs-Random conjecture. This is some evidence that AFC should not prevent the use of interactive classification for real-world tasks, as it is computationally hard to be exploited.

View on arXiv
Comments on this paper