# Video Question Answering on M3-Bench robot

Source: [Video Question Answering on M3-Bench robot](<https://www.sota2.com/research/sota/video-question-answering-on-m3-bench-robot>)

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

Data updated: 2026-07-07T07:22:38.565Z

Evaluation of model performance on the M3-Bench robot dataset for Video Question Answering.

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) | M3-Bench robot |
| Metric | ME Score |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 9 of 9 |
| 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

- [ME Score](<https://www.sota2.com/research/sota/video-question-answering-on-m3-bench-robot.md?metric=ME%20Score>): Higher is better; unit: SCALAR.

- [MH Score](<https://www.sota2.com/research/sota/video-question-answering-on-m3-bench-robot.md?metric=MH%20Score>): Higher is better; unit: SCALAR.

- [CM Score](<https://www.sota2.com/research/sota/video-question-answering-on-m3-bench-robot.md?metric=CM%20Score>): Higher is better; unit: SCALAR.

- [PU Score](<https://www.sota2.com/research/sota/video-question-answering-on-m3-bench-robot.md?metric=PU%20Score>): Higher is better; unit: SCALAR.

- [GK Score](<https://www.sota2.com/research/sota/video-question-answering-on-m3-bench-robot.md?metric=GK%20Score>): Higher is better; unit: SCALAR.

- [Overall Accuracy](<https://www.sota2.com/research/sota/video-question-answering-on-m3-bench-robot.md?metric=Overall%20Accuracy>): Higher is better; unit: SCALAR.

## Recorded results

| Method | ME Score | Result date | Report paper |
| --- | --- | --- | --- |
| HOMER | 43.7 | Not recorded | 6a4c87f385dba69fef8b2491 |
| M3-Agent-Gemini3 | 38.5 | Not recorded | 6a4c87f385dba69fef8b2491 |
| M3-Agent | 32.8 | Not recorded | 6a4c87f385dba69fef8b2491 |
| MA-LMM | 25.6 | Not recorded | 6a4c87f385dba69fef8b2491 |
| M3-Agent-Gemini25 | 24.6 | Not recorded | 6a4c87f385dba69fef8b2491 |
| Flash-VStream | 21.6 | Not recorded | 6a4c87f385dba69fef8b2491 |
| Gemini-GPT-Hybrid | 21.3 | Not recorded | 6a4c87f385dba69fef8b2491 |
| Gemini-Agent | 15.8 | Not recorded | 6a4c87f385dba69fef8b2491 |
| MovieChat | 13.3 | Not recorded | 6a4c87f385dba69fef8b2491 |

### HOMER

ME Score: 43.7. Result ID: 6a4ca92260b88bb339c7713b.

[Reporting paper](<https://www.sota2.com/research/paper/6a4c87f385dba69fef8b2491.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a4c87f385dba69fef8b2491.md>)

Recorded evaluation configuration: {"Method Category":"Agent Method","Backbone":"Gemini3"}. This record does not assert independent reproduction.

### M3-Agent-Gemini3

ME Score: 38.5. Result ID: 6a4ca92260b88bb339c7713a.

[Reporting paper](<https://www.sota2.com/research/paper/6a4c87f385dba69fef8b2491.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c021e209131cea8bd9c.md>)

Recorded evaluation configuration: {"Method Category":"Agent Method","Backbone":"Gemini3"}. This record does not assert independent reproduction.

### M3-Agent

ME Score: 32.8. Result ID: 6a4ca92260b88bb339c77138.

[Reporting paper](<https://www.sota2.com/research/paper/6a4c87f385dba69fef8b2491.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c021e209131cea8bd9c.md>)

Recorded evaluation configuration: {"Method Category":"Agent Method"}. This record does not assert independent reproduction.

### MA-LMM

ME Score: 25.6. Result ID: 6a4ca92260b88bb339c77134.

[Reporting paper](<https://www.sota2.com/research/paper/6a4c87f385dba69fef8b2491.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72f61e209131cea59960.md>)

[Implementation](<https://boheumd.github.io/MA-LMM/>)

Recorded evaluation configuration: {"Method Category":"Online Video Understanding"}. This record does not assert independent reproduction.

### M3-Agent-Gemini25

ME Score: 24.6. Result ID: 6a4ca92260b88bb339c77139.

[Reporting paper](<https://www.sota2.com/research/paper/6a4c87f385dba69fef8b2491.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c021e209131cea8bd9c.md>)

Recorded evaluation configuration: {"Method Category":"Agent Method","Backbone":"Gemini25"}. This record does not assert independent reproduction.

### Flash-VStream

ME Score: 21.6. Result ID: 6a4ca92260b88bb339c77135.

[Reporting paper](<https://www.sota2.com/research/paper/6a4c87f385dba69fef8b2491.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a291e209131cea81720.md>)

Recorded evaluation configuration: {"Method Category":"Online Video Understanding"}. This record does not assert independent reproduction.

### Gemini-GPT-Hybrid

ME Score: 21.3. Result ID: 6a4ca92260b88bb339c77137.

[Reporting paper](<https://www.sota2.com/research/paper/6a4c87f385dba69fef8b2491.md>)

Recorded evaluation configuration: {"Method Category":"Agent Method"}. This record does not assert independent reproduction.

### Gemini-Agent

ME Score: 15.8. Result ID: 6a4ca92260b88bb339c77136.

[Reporting paper](<https://www.sota2.com/research/paper/6a4c87f385dba69fef8b2491.md>)

Recorded evaluation configuration: {"Method Category":"Agent Method"}. This record does not assert independent reproduction.

### MovieChat

ME Score: 13.3. Result ID: 6a4ca92260b88bb339c77133.

[Reporting paper](<https://www.sota2.com/research/paper/6a4c87f385dba69fef8b2491.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72711e209131cea56563.md>)

[Implementation](<https://rese1f.github.io/MovieChat>)

Recorded evaluation configuration: {"Method Category":"Online Video Understanding"}. This record does not assert independent reproduction.

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