# GUI Agent Attack Success Rate Evaluation on MIRAGE (1,111-sample main set) Source: [GUI Agent Attack Success Rate Evaluation on MIRAGE (1,111-sample main set)]() Collection index: [llms.txt]() Data updated: 2026-05-28T09:01:57.034Z Evaluation of GUI Agent performance using Attack Success Rate metrics (FB, WA, Amaz, etc.) specifically on the 1,111-sample main set subset of the MIRAGE dataset. 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 | GUI Agent Attack Success Rate Evaluation | | Dataset (including any stated split) | MIRAGE (1,111-sample main set) | | Metric | FB Success Rate | | 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 - [FB Success Rate](): Higher is better; unit: PERCENT. - [WA Success Rate](): Higher is better; unit: PERCENT. - [Amaz Success Rate](): Higher is better; unit: PERCENT. - [IG Success Rate](): Higher is better; unit: PERCENT. - [Shop Success Rate](): Higher is better; unit: PERCENT. - [Spot Success Rate](): Higher is better; unit: PERCENT. - [Tel Success Rate](): Higher is better; unit: PERCENT. - [Temu Success Rate](): Higher is better; unit: PERCENT. - [TikT Success Rate](): Higher is better; unit: PERCENT. - [X Success Rate](): Higher is better; unit: PERCENT. - [Overall Success Rate](): Higher is better; unit: PERCENT. ## Recorded results | Method | FB Success Rate | Result date | Report paper | | --- | --- | --- | --- | | gpt-4o-mini | 41 | Not recorded | 6a17cc01dbd7150506f8123a | | Qwen3-VL-8B-Instruct | 29 | Not recorded | 6a17cc01dbd7150506f8123a | | GLM-4.5V | 28 | Not recorded | 6a17cc01dbd7150506f8123a | | Qwen3-VL-32B-Instruct | 27 | Not recorded | 6a17cc01dbd7150506f8123a | | Qwen3-VL-30B-A3B-Instruct | 26 | Not recorded | 6a17cc01dbd7150506f8123a | ### gpt-4o-mini FB Success Rate: 41. Result ID: 6a18032ffbc373592c11e243. [Reporting paper]() Recorded evaluation configuration: {"Model Type":"Closed-weight"}. This record does not assert independent reproduction. ### Qwen3-VL-8B-Instruct FB Success Rate: 29. Result ID: 6a18032ffbc373592c11e244. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"8B","Backbone":"Qwen3-VL"}. This record does not assert independent reproduction. ### GLM-4.5V FB Success Rate: 28. Result ID: 6a18032ffbc373592c11e245. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"GLM-4.5V"}. This record does not assert independent reproduction. ### Qwen3-VL-32B-Instruct FB Success Rate: 27. Result ID: 6a18032ffbc373592c11e247. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"32B","Backbone":"Qwen3-VL","Architecture":"Dense"}. This record does not assert independent reproduction. ### Qwen3-VL-30B-A3B-Instruct FB Success Rate: 26. Result ID: 6a18032ffbc373592c11e246. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"30B","Backbone":"Qwen3-VL","Architecture":"MoE"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.