# Video Question Answering on NExT-GQA (test)

Source: [Video Question Answering on NExT-GQA (test)](<https://www.sota2.com/research/sota/video-question-answering-on-next-gqa-test>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-02-26T05:26:58.090Z

Evaluation of model performance on the test split of the NExT-GQA dataset for Video Question Answering, reporting IoP, IoU, and GQA accuracy metrics.

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 (test) |
| Metric | Acc@GQA |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 29 of 29 |
| 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

- [Acc@GQA](<https://www.sota2.com/research/sota/video-question-answering-on-next-gqa-test.md?metric=Acc%40GQA>): Higher is better; unit: PERCENT.

- [mIoP](<https://www.sota2.com/research/sota/video-question-answering-on-next-gqa-test.md?metric=mIoP>): Higher is better; unit: PERCENT.

- [IoP@0.5](<https://www.sota2.com/research/sota/video-question-answering-on-next-gqa-test.md?metric=IoP%400.5>): Higher is better; unit: PERCENT.

- [mIoU](<https://www.sota2.com/research/sota/video-question-answering-on-next-gqa-test.md?metric=mIoU>): Higher is better; unit: PERCENT.

- [IoU@0.5](<https://www.sota2.com/research/sota/video-question-answering-on-next-gqa-test.md?metric=IoU%400.5>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Acc@GQA | Result date | Report paper |
| --- | --- | --- | --- |
| MoReVQA | 39.6 | Not recorded | 68fa764b1e209131cea6c30b |
| MoReVQA | 31 | Not recorded | 68fa77a41e209131cea748f6 |
| VITED (Temporal Evidence Distillation) | 27.61 | Not recorded | 68fa77751e209131cea733ca |
| VITED (Temporal Evidence Distillation) | 25.19 | Not recorded | 68fa77751e209131cea733ca |
| LLoVi | 24.3 | Not recorded | 68fa764b1e209131cea6c30b |
| FrozenBiLM | 17.5 | Not recorded | 68fa764b1e209131cea6c30b |
| FrozenBiLM | 17.5 | Not recorded | 68fa77a41e209131cea748f6 |
| LangRepo | 17.1 | Not recorded | 68fa764b1e209131cea6c30b |
| SeViLA | 16.6 | Not recorded | 68fa764b1e209131cea6c30b |
| SeViLA | 16.6 | Not recorded | 68fa77751e209131cea733ca |
| SeViLA | 16.6 | Not recorded | 68fa77a41e209131cea748f6 |
| LLoVi | 16.2 | Not recorded | 68fa764b1e209131cea6c30b |
| Temp\[CLIP\] | 16 | Not recorded | 68fa764b1e209131cea6c30b |
| Temp\[CLIP\] | 16 | Not recorded | 68fa77a41e209131cea748f6 |
| LLaMA-3.2V | 11.64 | Not recorded | 68fa77751e209131cea733ca |
| LLoVi | 11.2 | Not recorded | 68fa764b1e209131cea6c30b |
| LangRepo | 11.2 | Not recorded | 68fa764b1e209131cea6c30b |
| IGV | 10.2 | Not recorded | 68fa764b1e209131cea6c30b |
| IGV | 10.2 | Not recorded | 68fa77a41e209131cea748f6 |
| Mistral | 9.2 | Not recorded | 68fa764b1e209131cea6c30b |
| LLaVA-Video | 0.04 | Not recorded | 68fa77751e209131cea733ca |
| LLaVA-Video (Chain-of-Thought) | 0.03 | Not recorded | 68fa77751e209131cea733ca |
| LLaVA-OneVision | 0 | Not recorded | 68fa77751e209131cea733ca |
| TimeChat | 0 | Not recorded | 68fa77751e209131cea733ca |
| TimeChat (Chain-of-Thought) | 0 | Not recorded | 68fa77751e209131cea733ca |
| TimeChat (Video Instruction Tuning) | 0 | Not recorded | 68fa77751e209131cea733ca |
| VITED (Dense Caption Distillation) | 0 | Not recorded | 68fa77751e209131cea733ca |
| LLaVA-Video (Video Instruction Tuning) | 0 | Not recorded | 68fa77751e209131cea733ca |
| VITED (Dense Caption Distillation) | 0 | Not recorded | 68fa77751e209131cea733ca |

### MoReVQA

Acc@GQA: 39.6. Result ID: 699418813a6c3148491b059d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa764b1e209131cea6c30b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77a41e209131cea748f6.md>)

[Implementation](<http://juhongm999.github.io/morevqa>)

Recorded evaluation configuration: {"Video Pretrain":false,"Params":"340B","Protocol":"zero-shot (proprietary LLMs)"}. This record does not assert independent reproduction.

### MoReVQA

Acc@GQA: 31. Result ID: 69955a653badfaffe64baf9e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77a41e209131cea748f6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77a41e209131cea748f6.md>)

[Implementation](<http://juhongm999.github.io/morevqa>)

Recorded evaluation configuration: {"Fine-tuned":false}. This record does not assert independent reproduction.

### VITED (Temporal Evidence Distillation)

Acc@GQA: 27.61. Result ID: 69952ae03badfaffe610f63d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77751e209131cea733ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77751e209131cea733ca.md>)

Recorded evaluation configuration: {"Params":"7B","Base Model":"TimeChat Base"}. This record does not assert independent reproduction.

### VITED (Temporal Evidence Distillation)

Acc@GQA: 25.19. Result ID: 69952ae03badfaffe610f642.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77751e209131cea733ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77751e209131cea733ca.md>)

Recorded evaluation configuration: {"Params":"7B","Base Model":"LLaVA-Video Base"}. This record does not assert independent reproduction.

### LLoVi

Acc@GQA: 24.3. Result ID: 699418813a6c3148491b059e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa764b1e209131cea6c30b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75261e209131cea6629c.md>)

[Implementation](<https://github.com/CeeZh/LLoVi>)

Recorded evaluation configuration: {"Video Pretrain":false,"Params":"1.8T","Protocol":"zero-shot (proprietary LLMs)"}. This record does not assert independent reproduction.

### FrozenBiLM

Acc@GQA: 17.5. Result ID: 699418813a6c3148491b059b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa764b1e209131cea6c30b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6da21e209131cea3b08c.md>)

[Implementation](<https://antoyang.github.io/frozenbilm.html>)

Recorded evaluation configuration: {"Video Pretrain":true,"Params":"1B","Protocol":"weakly-supervised"}. This record does not assert independent reproduction.

### FrozenBiLM

Acc@GQA: 17.5. Result ID: 69955a653badfaffe64baf9c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77a41e209131cea748f6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6da21e209131cea3b08c.md>)

[Implementation](<https://antoyang.github.io/frozenbilm.html>)

Recorded evaluation configuration: {"Fine-tuned":true}. This record does not assert independent reproduction.

### LangRepo

Acc@GQA: 17.1. Result ID: 699418813a6c3148491b05a3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa764b1e209131cea6c30b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa764b1e209131cea6c30b.md>)

[Implementation](<https://www.sota2.com/github.com/kkahatapitiya/LangRepo>)

Recorded evaluation configuration: {"Video Pretrain":false,"Params":"12B","Protocol":"zero-shot (open-source LLMs)"}. This record does not assert independent reproduction.

### SeViLA

Acc@GQA: 16.6. Result ID: 699418813a6c3148491b059c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa764b1e209131cea6c30b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa715a1e209131cea50c7b.md>)

[Implementation](<https://github.com/Yui010206/SeViLA>)

Recorded evaluation configuration: {"Video Pretrain":true,"Params":"4B","Protocol":"weakly-supervised"}. This record does not assert independent reproduction.

### SeViLA

Acc@GQA: 16.6. Result ID: 69952ae03badfaffe610f636.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77751e209131cea733ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa715a1e209131cea50c7b.md>)

[Implementation](<https://github.com/Yui010206/SeViLA>)

Recorded evaluation configuration: {"Params":"4B"}. This record does not assert independent reproduction.

### SeViLA

Acc@GQA: 16.6. Result ID: 69955a653badfaffe64baf9d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77a41e209131cea748f6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa715a1e209131cea50c7b.md>)

[Implementation](<https://github.com/Yui010206/SeViLA>)

Recorded evaluation configuration: {"Fine-tuned":true,"Explicit localization supervision during pretraining":true}. This record does not assert independent reproduction.

### LLoVi

Acc@GQA: 16.2. Result ID: 699418813a6c3148491b05a1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa764b1e209131cea6c30b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75261e209131cea6629c.md>)

[Implementation](<https://github.com/CeeZh/LLoVi>)

Recorded evaluation configuration: {"Video Pretrain":false,"Params":"12B","Protocol":"zero-shot (open-source LLMs)"}. This record does not assert independent reproduction.

### Temp\[CLIP\]

Acc@GQA: 16. Result ID: 699418813a6c3148491b059a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa764b1e209131cea6c30b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69581e209131cea21ad0.md>)

[Implementation](<https://github.com/OpenAI/CLIP>)

Recorded evaluation configuration: {"Video Pretrain":true,"Params":"130M","Protocol":"weakly-supervised"}. This record does not assert independent reproduction.

### Temp\[CLIP\]

Acc@GQA: 16. Result ID: 69955a653badfaffe64baf9b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77a41e209131cea748f6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6da21e209131cea3b08c.md>)

[Implementation](<https://antoyang.github.io/frozenbilm.html>)

Recorded evaluation configuration: {"Fine-tuned":true}. This record does not assert independent reproduction.

### LLaMA-3.2V

Acc@GQA: 11.64. Result ID: 69952ae03badfaffe610f637.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77751e209131cea733ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d11e209131cea6a028.md>)

[Implementation](<https://github.com/meta-llama/llama-agentic-system>)

Recorded evaluation configuration: {"Params":"11B"}. This record does not assert independent reproduction.

### LLoVi

Acc@GQA: 11.2. Result ID: 699418813a6c3148491b05a0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa764b1e209131cea6c30b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75261e209131cea6629c.md>)

[Implementation](<https://github.com/CeeZh/LLoVi>)

Recorded evaluation configuration: {"Video Pretrain":false,"Params":"7B","Protocol":"zero-shot (open-source LLMs)"}. This record does not assert independent reproduction.

### LangRepo

Acc@GQA: 11.2. Result ID: 699418813a6c3148491b05a2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa764b1e209131cea6c30b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa764b1e209131cea6c30b.md>)

[Implementation](<https://www.sota2.com/github.com/kkahatapitiya/LangRepo>)

Recorded evaluation configuration: {"Video Pretrain":false,"Params":"7B","Protocol":"zero-shot (open-source LLMs)"}. This record does not assert independent reproduction.

### IGV

Acc@GQA: 10.2. Result ID: 699418813a6c3148491b0599.

[Reporting paper](<https://www.sota2.com/research/paper/68fa764b1e209131cea6c30b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6cbc1e209131cea35b17.md>)

Recorded evaluation configuration: {"Video Pretrain":true,"Params":"110M","Protocol":"weakly-supervised"}. This record does not assert independent reproduction.

### IGV

Acc@GQA: 10.2. Result ID: 69955a653badfaffe64baf9a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77a41e209131cea748f6.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6cbc1e209131cea35b17.md>)

Recorded evaluation configuration: {"Fine-tuned":true}. This record does not assert independent reproduction.

### Mistral

Acc@GQA: 9.2. Result ID: 699418813a6c3148491b059f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa764b1e209131cea6c30b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70d61e209131cea4d909.md>)

[Implementation](<https://github.com/mistralai/mistral-src>)

Recorded evaluation configuration: {"Video Pretrain":false,"Params":"7B","Protocol":"zero-shot (open-source LLMs)"}. This record does not assert independent reproduction.

### LLaVA-Video

Acc@GQA: 0.04. Result ID: 69952ae03badfaffe610f63e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77751e209131cea733ca.md>)

Recorded evaluation configuration: {"Params":"7B","Base Model":"LLaVA-Video Base"}. This record does not assert independent reproduction.

### LLaVA-Video (Chain-of-Thought)

Acc@GQA: 0.03. Result ID: 69952ae03badfaffe610f63f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77751e209131cea733ca.md>)

Recorded evaluation configuration: {"Params":"7B","Base Model":"LLaVA-Video Base"}. This record does not assert independent reproduction.

### LLaVA-OneVision

Acc@GQA: 0. Result ID: 69952ae03badfaffe610f638.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77751e209131cea733ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa756f1e209131cea67e58.md>)

[Implementation](<https://llava-vl.github.io/blog/llava-onevision>)

Recorded evaluation configuration: {"Params":"7B"}. This record does not assert independent reproduction.

### TimeChat

Acc@GQA: 0. Result ID: 69952ae03badfaffe610f639.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77751e209131cea733ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72ac1e209131cea57c30.md>)

[Implementation](<https://github.com/RenShuhuai-Andy/TimeChat>)

Recorded evaluation configuration: {"Params":"7B","Base Model":"TimeChat Base"}. This record does not assert independent reproduction.

### TimeChat (Chain-of-Thought)

Acc@GQA: 0. Result ID: 69952ae03badfaffe610f63a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77751e209131cea733ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72ac1e209131cea57c30.md>)

[Implementation](<https://github.com/RenShuhuai-Andy/TimeChat>)

Recorded evaluation configuration: {"Params":"7B","Base Model":"TimeChat Base"}. This record does not assert independent reproduction.

### TimeChat (Video Instruction Tuning)

Acc@GQA: 0. Result ID: 69952ae03badfaffe610f63b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77751e209131cea733ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72ac1e209131cea57c30.md>)

[Implementation](<https://github.com/RenShuhuai-Andy/TimeChat>)

Recorded evaluation configuration: {"Params":"7B","Base Model":"TimeChat Base"}. This record does not assert independent reproduction.

### VITED (Dense Caption Distillation)

Acc@GQA: 0. Result ID: 69952ae03badfaffe610f63c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77751e209131cea733ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77751e209131cea733ca.md>)

Recorded evaluation configuration: {"Params":"7B","Base Model":"TimeChat Base"}. This record does not assert independent reproduction.

### LLaVA-Video (Video Instruction Tuning)

Acc@GQA: 0. Result ID: 69952ae03badfaffe610f640.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77751e209131cea733ca.md>)

Recorded evaluation configuration: {"Params":"7B","Base Model":"LLaVA-Video Base"}. This record does not assert independent reproduction.

### VITED (Dense Caption Distillation)

Acc@GQA: 0. Result ID: 69952ae03badfaffe610f641.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77751e209131cea733ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77751e209131cea733ca.md>)

Recorded evaluation configuration: {"Params":"7B","Base Model":"LLaVA-Video Base"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.