arXiv:2506.03234v2 cs.LG

The Alignment Illusion: How BadReward Corrupts Text-to-Image Generation

Kaiwen Duan, Hongwei Yao, Yufei Chen, Ziyun Li, Tong Qiao, Zhan Qin, Cong Wang

제출 2025-06-03

원문 초록

Modern text-to-image generative models increasingly rely on Reinforcement Learning from Human Feedback (RLHF) to align generated content with human preferences and safety standards. This post-training stage is critical for optimizing models toward high-fidelity and value-aligned behaviors using human preference data. However, while previous works have extensively investigated poisoning attacks against the supervised fine-tuning (SFT) stage of text-to-image training, the security vulnerabilities introduced by the RLHF process remain unexplored. In this paper, we present the first exploration into the feasibility of poisoning the RLHF process as a novel attack surface for compromising text-to-image models. We introduce BadReward, an indirect clean-label poisoning attack that manipulates reward signals to embed malicious behaviors. Our approach leverages a feature collision mechanism to produce poison preference image pairs that are visually benign, but aligned with a malicious target concept in the reward feature space. Our experiments demonstrate that this manipulation effectively corrupts the reward signal. Online RLHF further amplifies this corruption through its reward-model feedback loop, whereas offline preference optimization learns directly from the poisoned preferences. Finally, we demonstrate the real-world effectiveness of BadReward on modern text-to-image architectures, showing high attack success rates and visual stealthiness across different RLHF algorithms. Our code is available at https://github.com/ZJUICSR/BadReward.

제출 이력

확인된 버전만 표시합니다. 전체 제출 이력은 arXiv 원문에서 확인하세요.

  1. arXiv:2506.03234v2제출 시각 미확인

함께 살펴볼 논문

공통 태그를 근거로 연결했습니다. 실제 인용 관계나 논문의 우수성을 뜻하지 않습니다.

공통 태그로 연결된 공개 논문이 아직 없습니다.

토론

이 논문과 연결된 포럼 글이 아직 없습니다.