# Multimodal Dialogue Response Generation on ImageChat seen prefixes (test)

Source: [Multimodal Dialogue Response Generation on ImageChat seen prefixes (test)](<https://www.sota2.com/research/sota/multimodal-dialogue-response-generation-on-imagechat-seen-prefixes-test>)

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

Data updated: 2026-02-26T06:06:01.092Z

Evaluation of multimodal dialogue response generation models using the test split of the ImageChat dataset, specifically focusing on responses conditioned on seen prefixes, measured by metrics like TR, Partial-F1, and TES.

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 | Multimodal Dialogue Response Generation |
| Dataset (including any stated split) | ImageChat seen prefixes (test) |
| Metric | TR |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 6 of 6 |
| 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

- [TR](<https://www.sota2.com/research/sota/multimodal-dialogue-response-generation-on-imagechat-seen-prefixes-test.md?metric=TR>): Higher is better; unit: SCALAR.

- [Syntactic Match](<https://www.sota2.com/research/sota/multimodal-dialogue-response-generation-on-imagechat-seen-prefixes-test.md?metric=Syntactic%20Match>): Higher is better; unit: PERCENT.

- [Precision (PR)](<https://www.sota2.com/research/sota/multimodal-dialogue-response-generation-on-imagechat-seen-prefixes-test.md?metric=Precision%20(PR)>): Higher is better; unit: PERCENT.

- [Recall (PR)](<https://www.sota2.com/research/sota/multimodal-dialogue-response-generation-on-imagechat-seen-prefixes-test.md?metric=Recall%20(PR)>): Higher is better; unit: PERCENT.

- [Partial-F1](<https://www.sota2.com/research/sota/multimodal-dialogue-response-generation-on-imagechat-seen-prefixes-test.md?metric=Partial-F1>): Higher is better; unit: PERCENT.

- [Avg Pred Len](<https://www.sota2.com/research/sota/multimodal-dialogue-response-generation-on-imagechat-seen-prefixes-test.md?metric=Avg%20Pred%20Len>): Higher is better; unit: SCALAR.

- [TES](<https://www.sota2.com/research/sota/multimodal-dialogue-response-generation-on-imagechat-seen-prefixes-test.md?metric=TES>): Lower is better; unit: SCALAR.

## Recorded results

| Method | TR | Result date | Report paper |
| --- | --- | --- | --- |
| MiniCPM-V | 0.9958 | Not recorded | 69647fda53fdd5c3617445a9 |
| Qwen2-VL | 0.9945 | Not recorded | 69647fda53fdd5c3617445a9 |
| PaliGemma | 0.9875 | Not recorded | 69647fda53fdd5c3617445a9 |
| QB | 0.9741 | Not recorded | 69647fda53fdd5c3617445a9 |
| MPC | 0.9497 | Not recorded | 69647fda53fdd5c3617445a9 |
| MPC++ | 0.9497 | Not recorded | 69647fda53fdd5c3617445a9 |

### MiniCPM-V

TR: 0.9958. Result ID: 699693995698665ec4dcd9eb.

[Reporting paper](<https://www.sota2.com/research/paper/69647fda53fdd5c3617445a9.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74681e209131cea6183d.md>)

[Implementation](<https://github.com/OpenBMB/MiniCPM-V>)

Recorded evaluation configuration: {"Model Type":"VLMs"}. This record does not assert independent reproduction.

### Qwen2-VL

TR: 0.9945. Result ID: 699693995698665ec4dcd9ed.

[Reporting paper](<https://www.sota2.com/research/paper/69647fda53fdd5c3617445a9.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75151e209131cea65976.md>)

[Implementation](<https://github.com/QwenLM/Qwen2-VL>)

Recorded evaluation configuration: {"Model Type":"VLMs"}. This record does not assert independent reproduction.

### PaliGemma

TR: 0.9875. Result ID: 699693995698665ec4dcd9ec.

[Reporting paper](<https://www.sota2.com/research/paper/69647fda53fdd5c3617445a9.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa751e1e209131cea65e0c.md>)

Recorded evaluation configuration: {"Model Type":"VLMs"}. This record does not assert independent reproduction.

### QB

TR: 0.9741. Result ID: 699693995698665ec4dcd9ea.

[Reporting paper](<https://www.sota2.com/research/paper/69647fda53fdd5c3617445a9.md>)

Recorded evaluation configuration: {"Model Type":"Text"}. This record does not assert independent reproduction.

### MPC

TR: 0.9497. Result ID: 699693995698665ec4dcd9e8.

[Reporting paper](<https://www.sota2.com/research/paper/69647fda53fdd5c3617445a9.md>)

Recorded evaluation configuration: {"Model Type":"Text"}. This record does not assert independent reproduction.

### MPC++

TR: 0.9497. Result ID: 699693995698665ec4dcd9e9.

[Reporting paper](<https://www.sota2.com/research/paper/69647fda53fdd5c3617445a9.md>)

Recorded evaluation configuration: {"Model Type":"Text"}. This record does not assert independent reproduction.

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