# Legal Moves Prediction on Chess Reasoning Dataset Legal Moves

Source: [Legal Moves Prediction on Chess Reasoning Dataset Legal Moves](<https://www.sota2.com/research/sota/legal-moves-prediction-on-chess-reasoning-dataset-legal-moves>)

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

Data updated: 2026-04-08T09:37:47.079Z

Evaluation of model performance on predicting the legality of chess moves using the Chess Reasoning Dataset Legal Moves, measured by IoU.

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 | Legal Moves Prediction |
| Dataset (including any stated split) | Chess Reasoning Dataset Legal Moves |
| Metric | IoU |
| Unit | PERCENT |
| 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

- [IoU](<https://www.sota2.com/research/sota/legal-moves-prediction-on-chess-reasoning-dataset-legal-moves.md?metric=IoU>): Higher is better; unit: PERCENT.

## Recorded results

| Method | IoU | Result date | Report paper |
| --- | --- | --- | --- |
| gpt-oss-120b (Medium) | 96 | Not recorded | 69d5e1267d0172140ec75b6b |
| SFT8 + SFT9 XL | 87 | Not recorded | 69d5e1267d0172140ec75b6b |
| SFT8 XL | 79 | Not recorded | 69d5e1267d0172140ec75b6b |
| SFT9 XL | 75 | Not recorded | 69d5e1267d0172140ec75b6b |
| SFT8 \[BM - All\] | 67 | Not recorded | 69d5e1267d0172140ec75b6b |
| SFT6 \[BM\] | 66 | Not recorded | 69d5e1267d0172140ec75b6b |
| SFT7 \[BL\] | 63 | Not recorded | 69d5e1267d0172140ec75b6b |
| SFT9 \[BL - All\] | 59 | Not recorded | 69d5e1267d0172140ec75b6b |
| SFT3 \[VABP\] | 58 | Not recorded | 69d5e1267d0172140ec75b6b |
| SFT4 \[FBA\] | 58 | Not recorded | 69d5e1267d0172140ec75b6b |
| SFT5 \[GS\] | 57 | Not recorded | 69d5e1267d0172140ec75b6b |
| SFT8 + SFT9 XL | 52 | Not recorded | 69d5e1267d0172140ec75b6b |
| SFT4 \[FBA\] | 49 | Not recorded | 69d5e1267d0172140ec75b6b |
| SFT9 XL | 47 | Not recorded | 69d5e1267d0172140ec75b6b |
| SFT7 \[BL\] | 46 | Not recorded | 69d5e1267d0172140ec75b6b |
| SFT8 XL | 46 | Not recorded | 69d5e1267d0172140ec75b6b |
| SFT2 \[RSA\] | 44 | Not recorded | 69d5e1267d0172140ec75b6b |
| SFT8 \[BM - All\] | 44 | Not recorded | 69d5e1267d0172140ec75b6b |
| Llama 4 Maverick | 43 | Not recorded | 69d5e1267d0172140ec75b6b |
| SFT6 \[BM\] | 42 | Not recorded | 69d5e1267d0172140ec75b6b |
| SFT3 \[VABP\] | 41 | Not recorded | 69d5e1267d0172140ec75b6b |
| SFT9 \[BL - All\] | 40 | Not recorded | 69d5e1267d0172140ec75b6b |
| SFT2 \[RSA\] | 37 | Not recorded | 69d5e1267d0172140ec75b6b |
| SFT5 \[GS\] | 37 | Not recorded | 69d5e1267d0172140ec75b6b |
| Qwen2.5 7B-Instruct | 26 | Not recorded | 69d5e1267d0172140ec75b6b |
| SFT1 \[RSPM\] | 26 | Not recorded | 69d5e1267d0172140ec75b6b |
| SFT1 \[RSPM\] | 17 | Not recorded | 69d5e1267d0172140ec75b6b |

### gpt-oss-120b (Medium)

IoU: 96. Result ID: 69d6213147f34ce924bf0b97.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

Recorded evaluation configuration: {"Training Stage":"Baseline"}. This record does not assert independent reproduction.

### SFT8 + SFT9 XL

IoU: 87. Result ID: 69d6213147f34ce924bf0baf.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

Recorded evaluation configuration: {"Training Stage":"RL"}. This record does not assert independent reproduction.

### SFT8 XL

IoU: 79. Result ID: 69d6213147f34ce924bf0bab.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

Recorded evaluation configuration: {"Training Stage":"RL"}. This record does not assert independent reproduction.

### SFT9 XL

IoU: 75. Result ID: 69d6213147f34ce924bf0bad.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

Recorded evaluation configuration: {"Training Stage":"RL"}. This record does not assert independent reproduction.

### SFT8 \[BM - All\]

IoU: 67. Result ID: 69d6213147f34ce924bf0ba7.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

Recorded evaluation configuration: {"Training Stage":"RL"}. This record does not assert independent reproduction.

### SFT6 \[BM\]

IoU: 66. Result ID: 69d6213147f34ce924bf0ba3.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

Recorded evaluation configuration: {"Training Stage":"RL"}. This record does not assert independent reproduction.

### SFT7 \[BL\]

IoU: 63. Result ID: 69d6213147f34ce924bf0ba5.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

Recorded evaluation configuration: {"Training Stage":"RL"}. This record does not assert independent reproduction.

### SFT9 \[BL - All\]

IoU: 59. Result ID: 69d6213147f34ce924bf0ba9.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

Recorded evaluation configuration: {"Training Stage":"RL"}. This record does not assert independent reproduction.

### SFT3 \[VABP\]

IoU: 58. Result ID: 69d6213147f34ce924bf0b9d.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

Recorded evaluation configuration: {"Training Stage":"RL"}. This record does not assert independent reproduction.

### SFT4 \[FBA\]

IoU: 58. Result ID: 69d6213147f34ce924bf0b9f.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

Recorded evaluation configuration: {"Training Stage":"RL"}. This record does not assert independent reproduction.

### SFT5 \[GS\]

IoU: 57. Result ID: 69d6213147f34ce924bf0ba1.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

Recorded evaluation configuration: {"Training Stage":"RL"}. This record does not assert independent reproduction.

### SFT8 + SFT9 XL

IoU: 52. Result ID: 69d6213147f34ce924bf0bae.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

Recorded evaluation configuration: {"Training Stage":"SFT"}. This record does not assert independent reproduction.

### SFT4 \[FBA\]

IoU: 49. Result ID: 69d6213147f34ce924bf0b9e.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

Recorded evaluation configuration: {"Training Stage":"SFT"}. This record does not assert independent reproduction.

### SFT9 XL

IoU: 47. Result ID: 69d6213147f34ce924bf0bac.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

Recorded evaluation configuration: {"Training Stage":"SFT"}. This record does not assert independent reproduction.

### SFT7 \[BL\]

IoU: 46. Result ID: 69d6213147f34ce924bf0ba4.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

Recorded evaluation configuration: {"Training Stage":"SFT"}. This record does not assert independent reproduction.

### SFT8 XL

IoU: 46. Result ID: 69d6213147f34ce924bf0baa.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

Recorded evaluation configuration: {"Training Stage":"SFT"}. This record does not assert independent reproduction.

### SFT2 \[RSA\]

IoU: 44. Result ID: 69d6213147f34ce924bf0b9b.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

Recorded evaluation configuration: {"Training Stage":"RL"}. This record does not assert independent reproduction.

### SFT8 \[BM - All\]

IoU: 44. Result ID: 69d6213147f34ce924bf0ba6.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

Recorded evaluation configuration: {"Training Stage":"SFT"}. This record does not assert independent reproduction.

### Llama 4 Maverick

IoU: 43. Result ID: 69d6213147f34ce924bf0b96.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

Recorded evaluation configuration: {"Training Stage":"Baseline"}. This record does not assert independent reproduction.

### SFT6 \[BM\]

IoU: 42. Result ID: 69d6213147f34ce924bf0ba2.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

Recorded evaluation configuration: {"Training Stage":"SFT"}. This record does not assert independent reproduction.

### SFT3 \[VABP\]

IoU: 41. Result ID: 69d6213147f34ce924bf0b9c.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

Recorded evaluation configuration: {"Training Stage":"SFT"}. This record does not assert independent reproduction.

### SFT9 \[BL - All\]

IoU: 40. Result ID: 69d6213147f34ce924bf0ba8.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

Recorded evaluation configuration: {"Training Stage":"SFT"}. This record does not assert independent reproduction.

### SFT2 \[RSA\]

IoU: 37. Result ID: 69d6213147f34ce924bf0b9a.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

Recorded evaluation configuration: {"Training Stage":"SFT"}. This record does not assert independent reproduction.

### SFT5 \[GS\]

IoU: 37. Result ID: 69d6213147f34ce924bf0ba0.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

Recorded evaluation configuration: {"Training Stage":"SFT"}. This record does not assert independent reproduction.

### Qwen2.5 7B-Instruct

IoU: 26. Result ID: 69d6213147f34ce924bf0b95.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

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

Recorded evaluation configuration: {"Training Stage":"Baseline"}. This record does not assert independent reproduction.

### SFT1 \[RSPM\]

IoU: 26. Result ID: 69d6213147f34ce924bf0b98.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

Recorded evaluation configuration: {"Training Stage":"SFT"}. This record does not assert independent reproduction.

### SFT1 \[RSPM\]

IoU: 17. Result ID: 69d6213147f34ce924bf0b99.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1267d0172140ec75b6b.md>)

Recorded evaluation configuration: {"Training Stage":"RL"}. This record does not assert independent reproduction.

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