# Science Question Answering on ARC Easy (Accuracy & Exit Position)

Source: [Science Question Answering on ARC Easy (Accuracy & Exit Position)](<https://www.sota2.com/research/sota/science-question-answering-on-arc-easy-accuracy-and-exit-position>)

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

Data updated: 2026-04-21T13:26:35.566Z

Evaluation of model performance on the Easy set of the Abstract Reasoning Corpus (ARC) for Science Question Answering, reporting Accuracy and Exit Position metrics.

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 | Science Question Answering |
| Dataset (including any stated split) | ARC Easy |
| Metric | 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

- [Accuracy](<https://www.sota2.com/research/sota/science-question-answering-on-arc-easy-accuracy-and-exit-position.md?metric=Accuracy>): Higher is better; unit: PERCENT.

- [Exit Position](<https://www.sota2.com/research/sota/science-question-answering-on-arc-easy-accuracy-and-exit-position.md?metric=Exit%20Position>): Lower is better; unit: SCALAR.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| Ministral3 8B | 86.5 | Not recorded | 69e705277d0172140ec7f309 |
| River-LLM | 85.9 | Not recorded | 69e705277d0172140ec7f309 |
| River-LLM | 85.9 | Not recorded | 69e705277d0172140ec7f309 |
| Phi4-mini | 82.8 | Not recorded | 69e705277d0172140ec7f309 |
| River-LLM | 81.3 | Not recorded | 69e705277d0172140ec7f309 |
| River-LLM | 81.3 | Not recorded | 69e705277d0172140ec7f309 |

### Ministral3 8B

Accuracy: 86.5. Result ID: 69e77ae847f34ce924cf62d5.

[Reporting paper](<https://www.sota2.com/research/paper/69e705277d0172140ec7f309.md>)

Recorded evaluation configuration: {"Model Architecture":"Ministral3 8B (34 layers)","Evaluation Protocol":"Backbone"}. This record does not assert independent reproduction.

### River-LLM

Accuracy: 85.9. Result ID: 69e77ae847f34ce924cf62d6.

[Reporting paper](<https://www.sota2.com/research/paper/69e705277d0172140ec7f309.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69e705277d0172140ec7f309.md>)

Recorded evaluation configuration: {"Model Architecture":"Ministral3 8B (34 layers)","Early-exit threshold (tau)":"0.8","Evaluation Protocol":"Adaptive"}. This record does not assert independent reproduction.

### River-LLM

Accuracy: 85.9. Result ID: 69e77ae847f34ce924cf62d7.

[Reporting paper](<https://www.sota2.com/research/paper/69e705277d0172140ec7f309.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69e705277d0172140ec7f309.md>)

Recorded evaluation configuration: {"Model Architecture":"Ministral3 8B (34 layers)","Early-exit threshold (tau)":"0.9","Evaluation Protocol":"Adaptive"}. This record does not assert independent reproduction.

### Phi4-mini

Accuracy: 82.8. Result ID: 69e77ae847f34ce924cf62d2.

[Reporting paper](<https://www.sota2.com/research/paper/69e705277d0172140ec7f309.md>)

Recorded evaluation configuration: {"Model Architecture":"Phi4-mini (32 layers)","Evaluation Protocol":"Backbone"}. This record does not assert independent reproduction.

### River-LLM

Accuracy: 81.3. Result ID: 69e77ae847f34ce924cf62d3.

[Reporting paper](<https://www.sota2.com/research/paper/69e705277d0172140ec7f309.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69e705277d0172140ec7f309.md>)

Recorded evaluation configuration: {"Model Architecture":"Phi4-mini (32 layers)","Early-exit threshold (tau)":"0.8","Evaluation Protocol":"Adaptive"}. This record does not assert independent reproduction.

### River-LLM

Accuracy: 81.3. Result ID: 69e77ae847f34ce924cf62d4.

[Reporting paper](<https://www.sota2.com/research/paper/69e705277d0172140ec7f309.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69e705277d0172140ec7f309.md>)

Recorded evaluation configuration: {"Model Architecture":"Phi4-mini (32 layers)","Early-exit threshold (tau)":"0.9","Evaluation Protocol":"Adaptive"}. This record does not assert independent reproduction.

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