# Multi-turn interaction accuracy on Multi-turn interaction follow-ups Round 3

Source: [Multi-turn interaction accuracy on Multi-turn interaction follow-ups Round 3](<https://www.sota2.com/research/sota/multi-turn-interaction-accuracy-on-multi-turn-interaction-follow-ups-round-3>)

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

Data updated: 2026-02-26T05:52:58.862Z

Evaluation of model performance on multi-turn conversational interactions using the Round 3 dataset, measured by interaction 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 | Multi-turn interaction accuracy |
| Dataset (including any stated split) | Multi-turn interaction follow-ups Round 3 |
| Metric | Multi-turn Interaction Accuracy |
| Unit | PERCENT |
| 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

- [Multi-turn Interaction Accuracy](<https://www.sota2.com/research/sota/multi-turn-interaction-accuracy-on-multi-turn-interaction-follow-ups-round-3.md?metric=Multi-turn%20Interaction%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Multi-turn Interaction Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| CARG | 74.29 | Not recorded | 68fa79d91e209131cea7fba0 |
| qwen | 67.29 | Not recorded | 68fa79d91e209131cea7fba0 |
| claude | 62 | Not recorded | 68fa79d91e209131cea7fba0 |
| mistral | 53.43 | Not recorded | 68fa79d91e209131cea7fba0 |
| llama | 48.43 | Not recorded | 68fa79d91e209131cea7fba0 |
| gemini | 41.29 | Not recorded | 68fa79d91e209131cea7fba0 |

### CARG

Multi-turn Interaction Accuracy: 74.29. Result ID: 698b628e7a967eddb76e28fd.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79d91e209131cea7fba0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa79d91e209131cea7fba0.md>)

[Implementation](<https://github.com/yubol-bobo/MT-Consistency>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### qwen

Multi-turn Interaction Accuracy: 67.29. Result ID: 698b628e7a967eddb76e28fc.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79d91e209131cea7fba0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa767e1e209131cea6d6a7.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### claude

Multi-turn Interaction Accuracy: 62. Result ID: 698b628e7a967eddb76e28f8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79d91e209131cea7fba0.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### mistral

Multi-turn Interaction Accuracy: 53.43. Result ID: 698b628e7a967eddb76e28f9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79d91e209131cea7fba0.md>)

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

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

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### llama

Multi-turn Interaction Accuracy: 48.43. Result ID: 698b628e7a967eddb76e28fa.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79d91e209131cea7fba0.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### gemini

Multi-turn Interaction Accuracy: 41.29. Result ID: 698b628e7a967eddb76e28fb.

[Reporting paper](<https://www.sota2.com/research/paper/68fa79d91e209131cea7fba0.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76361e209131cea6be11.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

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