# Zero-shot Image Classification on 15 Datasets (test) Source: [Zero-shot Image Classification on 15 Datasets (test)]() Collection index: [llms.txt]() Data updated: 2026-04-23T07:35:03.858Z Evaluation of Zero-shot Image Classification performance across a collection of 15 datasets, specifically reporting adversarial robustness metrics derived from PGD-20 attacks at various perturbation levels (epsilon 1/255 through 4/255) on the test set. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Zero-shot Image Classification | | Dataset (including any stated split) | 15 datasets Zero-shot (test) | | Metric | Robust Accuracy (PGD-20, eps=1/255) | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 5 of 5 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Robust Accuracy (PGD-20, eps=1/255)](): Higher is better; unit: PERCENT. - [Robust Accuracy (PGD-20, eps=2/255)](): Higher is better; unit: PERCENT. - [Robust Accuracy (PGD-20, eps=3/255)](): Higher is better; unit: PERCENT. - [Robust Accuracy (PGD-20, eps=4/255)](): Higher is better; unit: PERCENT. ## Recorded results | Method | Robust Accuracy (PGD-20, eps=1/255) | Result date | Report paper | | --- | --- | --- | --- | | Ours | 44.34 | Not recorded | 69e85bd47d0172140ec81123 | | AoS | 43.88 | Not recorded | 69e85bd47d0172140ec81123 | | PMG-FT | 39.72 | Not recorded | 69e85bd47d0172140ec81123 | | TeCoA | 38.91 | Not recorded | 69e85bd47d0172140ec81123 | | FARE | 37.93 | Not recorded | 69e85bd47d0172140ec81123 | ### Ours Robust Accuracy (PGD-20, eps=1/255): 44.34. Result ID: 69e9cb8a47f34ce924d09dfa. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Attack protocol":"PGD-20","Evaluation mode":"Zero-shot"}. This record does not assert independent reproduction. ### AoS Robust Accuracy (PGD-20, eps=1/255): 43.88. Result ID: 69e9cb8a47f34ce924d09df9. [Reporting paper]() Recorded evaluation configuration: {"Attack protocol":"PGD-20","Evaluation mode":"Zero-shot"}. This record does not assert independent reproduction. ### PMG-FT Robust Accuracy (PGD-20, eps=1/255): 39.72. Result ID: 69e9cb8a47f34ce924d09df7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Attack protocol":"PGD-20","Evaluation mode":"Zero-shot"}. This record does not assert independent reproduction. ### TeCoA Robust Accuracy (PGD-20, eps=1/255): 38.91. Result ID: 69e9cb8a47f34ce924d09df6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Attack protocol":"PGD-20","Evaluation mode":"Zero-shot"}. This record does not assert independent reproduction. ### FARE Robust Accuracy (PGD-20, eps=1/255): 37.93. Result ID: 69e9cb8a47f34ce924d09df8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Attack protocol":"PGD-20","Evaluation mode":"Zero-shot"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.