# Video Question Answering on NExT-GQA Source: [Video Question Answering on NExT-GQA]() Collection index: [llms.txt]() Data updated: 2026-04-07T10:22:01.714Z Evaluation of model performance on the NExT-GQA dataset for Video Question Answering, reporting Accuracy. 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 | Video Question Answering | | Dataset (including any stated split) | NExT-GQA | | Metric | Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 13 of 13 | | 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](): Higher is better; unit: PERCENT. ## Recorded results | Method | Accuracy | Result date | Report paper | | --- | --- | --- | --- | | SDRL | 79.3 | Not recorded | 69ca039e0636e5bf4afc942a | | SDRL | 77.3 | Not recorded | 69ca039e0636e5bf4afc942a | | TW-GRPO | 76.1 | Not recorded | 69ca039e0636e5bf4afc942a | | VideoChat-R1 | 76 | Not recorded | 69ca039e0636e5bf4afc942a | | Qwen2.5-VL-7B | 75.9 | Not recorded | 69ca039e0636e5bf4afc942a | | VideoRFT | 75.1 | Not recorded | 69ca039e0636e5bf4afc942a | | Video-R1 | 74.3 | Not recorded | 69ca039e0636e5bf4afc942a | | Qwen2.5-VL-7B | 73.6 | Not recorded | 69ca039e0636e5bf4afc942a | | Factum-4B | 73.6 | Not recorded | 69d48faf7d0172140ec74469 | | Qwen3-VL-4B-Instruct | 72.1 | Not recorded | 69d48faf7d0172140ec74469 | | Qwen3-VL-4B-Thinking | 66.6 | Not recorded | 69d48faf7d0172140ec74469 | | Qwen2.5-VL-7B | 59.5 | Not recorded | 69d48faf7d0172140ec74469 | | VTimeLLM-7B | 17.4 | Not recorded | 69d48faf7d0172140ec74469 | ### SDRL Accuracy: 79.3. Result ID: 69ca1d9b47f34ce924b22f22. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training":"RL"}. This record does not assert independent reproduction. ### SDRL Accuracy: 77.3. Result ID: 69ca1d9b47f34ce924b22f21. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training":"RL","Training dataset":"EventFlow"}. This record does not assert independent reproduction. ### TW-GRPO Accuracy: 76.1. Result ID: 69ca1d9b47f34ce924b22f20. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training":"RL"}. This record does not assert independent reproduction. ### VideoChat-R1 Accuracy: 76. Result ID: 69ca1d9b47f34ce924b22f1f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training":"RL"}. This record does not assert independent reproduction. ### Qwen2.5-VL-7B Accuracy: 75.9. Result ID: 69ca1d9b47f34ce924b22f1b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training":"None"}. This record does not assert independent reproduction. ### VideoRFT Accuracy: 75.1. Result ID: 69ca1d9b47f34ce924b22f1e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training":"SFT+ RL","Input frames":"16-frame"}. This record does not assert independent reproduction. ### Video-R1 Accuracy: 74.3. Result ID: 69ca1d9b47f34ce924b22f1d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training":"SFT+ RL"}. This record does not assert independent reproduction. ### Qwen2.5-VL-7B Accuracy: 73.6. Result ID: 69ca1d9b47f34ce924b22f1c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Training":"None","Chain-of-Thought (CoT)":"ours CoT"}. This record does not assert independent reproduction. ### Factum-4B Accuracy: 73.6. Result ID: 69d4dac647f34ce924bd40c3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Source Type":"Open-Source","FPS":"1fps"}. This record does not assert independent reproduction. ### Qwen3-VL-4B-Instruct Accuracy: 72.1. Result ID: 69d4dac647f34ce924bd40c1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Source Type":"Open-Source","FPS":"1fps"}. This record does not assert independent reproduction. ### Qwen3-VL-4B-Thinking Accuracy: 66.6. Result ID: 69d4dac647f34ce924bd40c2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Source Type":"Open-Source","FPS":"1fps"}. This record does not assert independent reproduction. ### Qwen2.5-VL-7B Accuracy: 59.5. Result ID: 69d4dac647f34ce924bd40c0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Source Type":"Open-Source","FPS":"1fps"}. This record does not assert independent reproduction. ### VTimeLLM-7B Accuracy: 17.4. Result ID: 69d4dac647f34ce924bd40bf. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Source Type":"Open-Source","FPS":"1fps"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.