# Visual Question Answering on GQA (Epsilon Robustness Accuracy) Source: [Visual Question Answering on GQA (Epsilon Robustness Accuracy)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:34:31.349Z Evaluation of model performance on the GQA dataset for Visual Question Answering, focusing specifically on accuracy under different levels of input perturbation (epsilon values). 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 | Visual Question Answering | | Dataset (including any stated split) | GQA | | Metric | Accuracy (clean) | | 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 - [Accuracy (clean)](): Higher is better; unit: PERCENT. - [Accuracy (eps=2/255)](): Higher is better; unit: PERCENT. - [Accuracy (eps=4/255)](): Higher is better; unit: PERCENT. ## Recorded results | Method | Accuracy (clean) | Result date | Report paper | | --- | --- | --- | --- | | UniTok-MLLM | 68 | Not recorded | 699bdef109c33900fcd8041f | | UniTok-MLLM | 67 | Not recorded | 699bdef109c33900fcd8041f | | UniTok-MLLM | 65.4 | Not recorded | 699bdef109c33900fcd8041f | | UniTok-MLLM | 64.2 | Not recorded | 699bdef109c33900fcd8041f | | UniTok-MLLM | 61.2 | Not recorded | 699bdef109c33900fcd8041f | ### UniTok-MLLM Accuracy (clean): 68. Result ID: 699d09a65698665ec4a9ad70. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tokenizer":"clean"}. This record does not assert independent reproduction. ### UniTok-MLLM Accuracy (clean): 67. Result ID: 699d09a65698665ec4a9ad71. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tokenizer":"AT 4/255"}. This record does not assert independent reproduction. ### UniTok-MLLM Accuracy (clean): 65.4. Result ID: 699d09a65698665ec4a9ad72. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tokenizer":"AT 8/255"}. This record does not assert independent reproduction. ### UniTok-MLLM Accuracy (clean): 64.2. Result ID: 699d09a65698665ec4a9ad73. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tokenizer":"AT 12/255"}. This record does not assert independent reproduction. ### UniTok-MLLM Accuracy (clean): 61.2. Result ID: 699d09a65698665ec4a9ad74. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Tokenizer":"AT 16/255"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.