# Science Question Answering on AI2ARC (test)

Source: [Science Question Answering on AI2ARC (test)](<https://www.sota2.com/research/sota/science-question-answering-on-ai2arc-test>)

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

Data updated: 2026-02-26T05:29:29.390Z

Evaluation of model performance on the test split of the AI2ARC dataset for Science Question Answering.

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) | AI2ARC (test) |
| 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-ai2arc-test.md?metric=Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| GPT-4 | 96.3 | Not recorded | 68fa727a1e209131cea56a1e |
| ST-MoE | 86.5 | Not recorded | 68fa727a1e209131cea56a1e |
| GPT-3.5 | 85.2 | Not recorded | 68fa727a1e209131cea56a1e |
| PaLM | 85.2 | Not recorded | 68fa727a1e209131cea56a1e |
| TAMA | 81.23 | Not recorded | 68fa7a501e209131cea82659 |
| base | 80.89 | Not recorded | 68fa7a501e209131cea82659 |

### GPT-4

Accuracy: 96.3. Result ID: 69923ce73a6c3148490d85fe.

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

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

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

Recorded evaluation configuration: {"number\_of\_shots":"25-shot"}. This record does not assert independent reproduction.

### ST-MoE

Accuracy: 86.5. Result ID: 69923ce73a6c3148490d8601.

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

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c641e209131cea33c61.md>)

[Implementation](<https://github.com/tensorflow/mesh/blob/master/mesh_tensorflow/transformer/moe.py>)

Recorded evaluation configuration: {"benchmark-specific tuning":true}. This record does not assert independent reproduction.

### GPT-3.5

Accuracy: 85.2. Result ID: 69923ce73a6c3148490d85ff.

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

Recorded evaluation configuration: {"number\_of\_shots":"25-shot"}. This record does not assert independent reproduction.

### PaLM

Accuracy: 85.2. Result ID: 69923ce73a6c3148490d8600.

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

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d931e209131cea3abf2.md>)

Recorded evaluation configuration: {"number\_of\_shots":"8-shot"}. This record does not assert independent reproduction.

### TAMA

Accuracy: 81.23. Result ID: 698b74487a967eddb76eb604.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a501e209131cea82659.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a501e209131cea82659.md>)

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

### base

Accuracy: 80.89. Result ID: 698b74487a967eddb76eb603.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a501e209131cea82659.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75d11e209131cea6a028.md>)

[Implementation](<https://github.com/meta-llama/llama-agentic-system>)

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.