# EDA on Celebrities-30 EDA (subsampled)

Source: [EDA on Celebrities-30 EDA (subsampled)](<https://www.sota2.com/research/sota/eda-on-celebrities-30-eda-subsampled>)

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

Data updated: 2026-02-26T05:21:18.665Z

Evaluation of model performance on the Celebrities-30 EDA (subsampled) dataset, measured using interaction metrics such as #Turns, Success Rate, and Score.

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 | EDA |
| Dataset (including any stated split) | Celebrities-30 EDA (subsampled) |
| Metric | #Turns |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 27 of 27 |
| 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

- [#Turns](<https://www.sota2.com/research/sota/eda-on-celebrities-30-eda-subsampled.md?metric=%23Turns>): Higher is better; unit: SCALAR.

- [Success Rate](<https://www.sota2.com/research/sota/eda-on-celebrities-30-eda-subsampled.md?metric=Success%20Rate>): Higher is better; unit: PERCENT.

- [#Yes](<https://www.sota2.com/research/sota/eda-on-celebrities-30-eda-subsampled.md?metric=%23Yes>): Higher is better; unit: SCALAR.

- [Score](<https://www.sota2.com/research/sota/eda-on-celebrities-30-eda-subsampled.md?metric=Score>): Higher is better; unit: SCALAR.

## Recorded results

| Method | #Turns | Result date | Report paper |
| --- | --- | --- | --- |
| Vicuna | 19.7 | Not recorded | 68fa72471e209131cea55847 |
| Vicuna 7B | 19.6 | Not recorded | 68fa72471e209131cea55847 |
| V-RLGP | 19.5 | Not recorded | 68fa72471e209131cea55847 |
| V-FT 7B (All) | 19.3 | Not recorded | 68fa72471e209131cea55847 |
| V-FT | 19.1 | Not recorded | 68fa72471e209131cea55847 |
| V-FT | 19.1 | Not recorded | 68fa72471e209131cea55847 |
| V-FT | 19 | Not recorded | 68fa72471e209131cea55847 |
| V-FT 7B (Suc.) | 19 | Not recorded | 68fa72471e209131cea55847 |
| GPT-3.5 | 18.8 | Not recorded | 68fa72471e209131cea55847 |
| V-FT 13B (All) | 18.8 | Not recorded | 68fa72471e209131cea55847 |
| V-RLGP 7B | 18.8 | Not recorded | 68fa72471e209131cea55847 |
| Vicuna 13B | 18.7 | Not recorded | 68fa72471e209131cea55847 |
| V-FT | 18.6 | Not recorded | 68fa72471e209131cea55847 |
| V-FT | 18.6 | Not recorded | 68fa72471e209131cea55847 |
| V-RLGP 13B | 18.5 | Not recorded | 68fa72471e209131cea55847 |
| V-FT 13B (Suc.) | 18.4 | Not recorded | 68fa72471e209131cea55847 |
| Mistral 7B | 18.2 | Not recorded | 68fa72471e209131cea55847 |
| Human | 18.1 | Not recorded | 68fa72471e209131cea55847 |
| GPT-3.5 | 17.9 | Not recorded | 68fa72471e209131cea55847 |
| Vicuna | 17.7 | Not recorded | 68fa72471e209131cea55847 |
| Claude-1 | 17.7 | Not recorded | 68fa72471e209131cea55847 |
| Claude-2 | 17.6 | Not recorded | 68fa72471e209131cea55847 |
| V-RLGP | 17.5 | Not recorded | 68fa72471e209131cea55847 |
| GPT-4 | 17.3 | Not recorded | 68fa72471e209131cea55847 |
| Claude-1 | 16.7 | Not recorded | 68fa72471e209131cea55847 |
| GPT-4 | 16.5 | Not recorded | 68fa72471e209131cea55847 |
| Claude-2 | 15.9 | Not recorded | 68fa72471e209131cea55847 |

### Vicuna

#Turns: 19.7. Result ID: 69922ac13a6c3148490d08cd.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

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

### Vicuna 7B

#Turns: 19.6. Result ID: 69922ac13a6c3148490d08e8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

Recorded evaluation configuration: {"Size":"7B","Version":"v1.3"}. This record does not assert independent reproduction.

### V-RLGP

#Turns: 19.5. Result ID: 69922ac13a6c3148490d08d3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

[Implementation](<https://github.com/apple/ml-entity-deduction-arena>)

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

### V-FT 7B (All)

#Turns: 19.3. Result ID: 69922ac13a6c3148490d08ea.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

[Implementation](<https://github.com/apple/ml-entity-deduction-arena>)

Recorded evaluation configuration: {"Size":"7B","Training Data":"All turns"}. This record does not assert independent reproduction.

### V-FT

#Turns: 19.1. Result ID: 69922ac13a6c3148490d08cf.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

[Implementation](<https://github.com/apple/ml-entity-deduction-arena>)

Recorded evaluation configuration: {"Backbone":"Vicuna 7B","Training Data":"Things"}. This record does not assert independent reproduction.

### V-FT

#Turns: 19.1. Result ID: 69922ac13a6c3148490d08d0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

[Implementation](<https://github.com/apple/ml-entity-deduction-arena>)

Recorded evaluation configuration: {"Backbone":"Vicuna 7B","Training Data":"Celebs"}. This record does not assert independent reproduction.

### V-FT

#Turns: 19. Result ID: 69922ac13a6c3148490d08ce.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

[Implementation](<https://github.com/apple/ml-entity-deduction-arena>)

Recorded evaluation configuration: {"Backbone":"Vicuna 7B","Training Data":"All"}. This record does not assert independent reproduction.

### V-FT 7B (Suc.)

#Turns: 19. Result ID: 69922ac13a6c3148490d08eb.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

[Implementation](<https://github.com/apple/ml-entity-deduction-arena>)

Recorded evaluation configuration: {"Size":"7B","Training Data":"Successful turns only"}. This record does not assert independent reproduction.

### GPT-3.5

#Turns: 18.8. Result ID: 69922ac13a6c3148490d08e4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

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

### V-FT 13B (All)

#Turns: 18.8. Result ID: 69922ac13a6c3148490d08ec.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

[Implementation](<https://github.com/apple/ml-entity-deduction-arena>)

Recorded evaluation configuration: {"Size":"13B","Training Data":"All turns"}. This record does not assert independent reproduction.

### V-RLGP 7B

#Turns: 18.8. Result ID: 69922ac13a6c3148490d08ee.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

[Implementation](<https://github.com/apple/ml-entity-deduction-arena>)

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

### Vicuna 13B

#Turns: 18.7. Result ID: 69922ac13a6c3148490d08e7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

Recorded evaluation configuration: {"Size":"13B","Version":"v1.3"}. This record does not assert independent reproduction.

### V-FT

#Turns: 18.6. Result ID: 69922ac13a6c3148490d08d1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

[Implementation](<https://github.com/apple/ml-entity-deduction-arena>)

Recorded evaluation configuration: {"Backbone":"Vicuna 7B","Training Data":"Success trajectories"}. This record does not assert independent reproduction.

### V-FT

#Turns: 18.6. Result ID: 69922ac13a6c3148490d08d2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

[Implementation](<https://github.com/apple/ml-entity-deduction-arena>)

Recorded evaluation configuration: {"Backbone":"Vicuna 13B","Training Data":"Success trajectories"}. This record does not assert independent reproduction.

### V-RLGP 13B

#Turns: 18.5. Result ID: 69922ac13a6c3148490d08ef.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

[Implementation](<https://github.com/apple/ml-entity-deduction-arena>)

Recorded evaluation configuration: {"Size":"13B"}. This record does not assert independent reproduction.

### V-FT 13B (Suc.)

#Turns: 18.4. Result ID: 69922ac13a6c3148490d08ed.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

[Implementation](<https://github.com/apple/ml-entity-deduction-arena>)

Recorded evaluation configuration: {"Size":"13B","Training Data":"Successful turns only"}. This record does not assert independent reproduction.

### Mistral 7B

#Turns: 18.2. Result ID: 69922ac13a6c3148490d08e9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

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

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

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

### Human

#Turns: 18.1. Result ID: 69922ac13a6c3148490d08d5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

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

### GPT-3.5

#Turns: 17.9. Result ID: 69922ac13a6c3148490d08c9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

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

### Vicuna

#Turns: 17.7. Result ID: 69922ac13a6c3148490d08cc.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

Recorded evaluation configuration: {"Model size":"13B"}. This record does not assert independent reproduction.

### Claude-1

#Turns: 17.7. Result ID: 69922ac13a6c3148490d08e6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

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

### Claude-2

#Turns: 17.6. Result ID: 69922ac13a6c3148490d08e5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

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

### V-RLGP

#Turns: 17.5. Result ID: 69922ac13a6c3148490d08d4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

[Implementation](<https://github.com/apple/ml-entity-deduction-arena>)

Recorded evaluation configuration: {"Backbone":"Vicuna 13B"}. This record does not assert independent reproduction.

### GPT-4

#Turns: 17.3. Result ID: 69922ac13a6c3148490d08e3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa727a1e209131cea56a1e.md>)

[Implementation](<https://github.com/openai/evals>)

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

### Claude-1

#Turns: 16.7. Result ID: 69922ac13a6c3148490d08cb.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

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

### GPT-4

#Turns: 16.5. Result ID: 69922ac13a6c3148490d08c8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa727a1e209131cea56a1e.md>)

[Implementation](<https://github.com/openai/evals>)

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

### Claude-2

#Turns: 15.9. Result ID: 69922ac13a6c3148490d08ca.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72471e209131cea55847.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.
