A novel black-box image embedding inversion attack that reconstructs high-fidelity images using only query access to the embedding model or API, and introduces an embedding-guided cross-attention mechanism, where image embeddings serve as conditional signals to steer the generation process.
Lichao Sun, Yun-Cheng Wu, Hai-Chao Sha et al.· Proceedings of the 32nd ACM...· 0 citations
This paper proposes Disentangled Action--Reasoning Tuning (DART), a simple and efficient framework that explicitly decouples parameter updates for reasoning and tool use via separate low-rank adaptation modules, further supporting the finding of capability interference under shared optimization.
Yu Li, Ming-Yang Yi, Xiu‐Qing Li et al.· arXiv.org· 15 citations· ⚡1
A novel black-box image embedding inversion attack that reconstructs high-fidelity images using only query access to the embedding model or API, and introduces an embedding-guided cross-attention mechanism, where image embeddings serve as conditional signals to steer the generation process.
Lichao Sun, Yuncheng Wu, Haichao Sha et al.· Proceedings of the 32nd ACM...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
