Newest papers

Adversarial Benchmarks and BenchDoor Attacks: Exploiting Benchmark Incentives for Backdoor Attacks

Yixin Wu, Rachel S. Hong, Jamie Morgenstern, Tadayoshi Kohno · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.27533 · cited by 0 works (OpenAlex) · open-access PDF

Korean

기계 학습 벤치마크는 모델 개발을위한 데이터 가이드 포스트로 작용하고 벤치마크 점수를 향상시키는 훈련 데이터 세트의 창출을 촉진합니다. 우리는 이 자극을 활용하는 데이터 중독 공격의 가족인 BenchDoor를 소개합니다. 벤치마크 성능을 보존하는 것 외에도 BenchDoor는 벤치마크 점수를 지원하는 데이터의 채택을 촉진합니다. 상대방은 벤치마크를 사용하여 벤치마크를 사용하여 벤치마크 벤치마크를 사용하여 벤치마크 벤치마크를 사용하여 벤치마크 벤치마크를 사용하여 벤치마크 벤치마크 데이터 세트를 제공합니다. BenchDoor는 벤치마크를

Machine translation (Meta M2M100); the English original follows.

Abstract

Machine learning benchmarks act as guideposts for model development and spur the creation of training datasets intended to improve benchmark scores. We introduce BenchDoor, a family of data poisoning attacks that exploits this incentive. Beyond preserving benign performance, BenchDoor requires benchmark gains that encourage adoption of data that also implants a backdoor. The adversary either manipulates a benchmark's training or test data, creating an adversarial benchmark, or supplies an independent training dataset while leaving the benchmark unchanged. We formalize five attack scenarios, including training on rewritten test examples and applying similarity-based decontamination. Across multiple-choice Q&A, text classification, and open-ended Q&A with Qwen and Gemma models, BenchDoor achieves positive benchmark gains for all 30 evaluated configurations, with held-out attack success rates of 82.33% to 100.00% at poisoning rates of 2% to 10%. Both objectives remain achievable after synthetic data generation and similarity-based decontamination, with attack success rates of at least 90.33%. These findings show how benchmarks can serve as attack vectors both through their own data and through training resources that deliver improvements on them.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Benchmark (surveying), Backdoor, Exploit, Adversarial system, Computer science, Adversary

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange