PROVE：基于可验证证据的免训练提示词恢复方法
原标题：PROVE: Training-Free Prompt Recovery using Verifiable Evidence
AI 摘要
PROVE 是一种无需训练的、黑盒的提示词恢复攻击方法，通过组合可验证的场景描述而非优化 token 序列来重建提示词，适用于原始版权作品和 AI 生成内容。在 MS-COCO、Flickr30K 和 Lexica 数据集上，PROVE 在图像相似度和文本-图像对齐方面优于基于优化、字幕和强化学习的基线方法，且无需训练或访问生成器。该方法旨在应对提示词市场兴起带来的版权保护和内容所有权风险，同时提升模型透明度和可解释性。
正文节选
PROVE: Training-Free Prompt Recovery using Verifiable Evidence Abstract Modern text-to-image models can generate highly realistic images from natural-language prompts, while recent advances in prompt inversion have made it increasingly feasible to recover those prompts from generated outputs, raising new concerns for copyright protection and content ownership. As prompt marketplaces emerge, recovered prompts can enable both the unauthorized reproduction and redistribution of copyrighted creative