Boosting the visual interpretability of CLIP via adversarial fine-tuning

Shizhan Gong, Haoyu Lei, Qi Dou, Farzan Farnia

2/12/2026

No official code found. Best implementation: peterant330/CLIP_AFT (11 stars, last commit 2025-03-10).

Semantic Scholar

Code Implementations(1)

peterant330/CLIP_AFTCommunity63%

[ICLR'25] Boosting the visual interpretability of CLIP via adversarial fine-tuning

110Python, Jupyter NotebookFeb 19, 20251 years ago

Cite this paper

@article{gong2026boosting,
  title  = {Boosting the visual interpretability of CLIP via adversarial fine-tuning},
  author = {Shizhan Gong and Haoyu Lei and Qi Dou and Farzan Farnia},
  year   = {2026},
  url    = {https://api.semanticscholar.org/CorpusID:a6cdeb6106e1be5787dac03412812e115455fcea},
  journal = {ICLR 2025 2025}
}

Boosting the visual interpretability of CLIP via adversarial fine-tuning

Code Implementations(1)

Cite this paper

Discussion