English

Low-Loss Subspace Compression for Clean Gains against Multi-Agent Backdoor Attacks

Machine Learning 2022-09-21 v2 Cryptography and Security Multiagent Systems

Abstract

Recent exploration of the multi-agent backdoor attack demonstrated the backfiring effect, a natural defense against backdoor attacks where backdoored inputs are randomly classified. This yields a side-effect of low accuracy w.r.t. clean labels, which motivates this paper's work on the construction of multi-agent backdoor defenses that maximize accuracy w.r.t. clean labels and minimize that of poison labels. Founded upon agent dynamics and low-loss subspace construction, we contribute three defenses that yield improved multi-agent backdoor robustness.

Cite

@article{arxiv.2203.03692,
  title  = {Low-Loss Subspace Compression for Clean Gains against Multi-Agent Backdoor Attacks},
  author = {Siddhartha Datta and Nigel Shadbolt},
  journal= {arXiv preprint arXiv:2203.03692},
  year   = {2022}
}
R2 v1 2026-06-24T10:05:12.306Z