# Science Question Answering on ARC (Easy/Challenge Breakdown)

Source: [Science Question Answering on ARC (Easy/Challenge Breakdown)](<https://www.sota2.com/research/sota/science-question-answering-on-arc-easy-challenge-breakdown>)

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

Data updated: 2026-02-26T05:23:37.956Z

Evaluation of model performance on the Abstract Reasoning Corpus (ARC) dataset, reporting separate accuracy scores for the Easy and Challenge subsets.

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 |
| Metric | ARC Easy Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 16 of 16 |
| 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

- [ARC Easy Accuracy](<https://www.sota2.com/research/sota/science-question-answering-on-arc-easy-challenge-breakdown.md?metric=ARC%20Easy%20Accuracy>): Higher is better; unit: PERCENT.

- [ARC Challenge Accuracy](<https://www.sota2.com/research/sota/science-question-answering-on-arc-easy-challenge-breakdown.md?metric=ARC%20Challenge%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | ARC Easy Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| LD-MoLE | 92.11 | Not recorded | 68fa7bc81e209131cea8a6fb |
| ReMoLE | 91.75 | Not recorded | 68fa7bc81e209131cea8a6fb |
| ReMoLE | 91.61 | Not recorded | 68fa7bc81e209131cea8a6fb |
| LD-MoLE | 91.61 | Not recorded | 68fa7bc81e209131cea8a6fb |
| LD-MoLE | 89.47 | Not recorded | 68fa7bc81e209131cea8a6fb |
| ReMoLE | 89.3 | Not recorded | 68fa7bc81e209131cea8a6fb |
| MoLA(8888) | 88.6 | Not recorded | 68fa7bc81e209131cea8a6fb |
| MoLA(2468) | 88.42 | Not recorded | 68fa7bc81e209131cea8a6fb |
| MoLA(2468) | 87.37 | Not recorded | 68fa7bc81e209131cea8a6fb |
| MoLA(8888) | 86.32 | Not recorded | 68fa7bc81e209131cea8a6fb |
| MoLA(2468) | 83.86 | Not recorded | 68fa7bc81e209131cea8a6fb |
| MoLA(8888) | 83.51 | Not recorded | 68fa7bc81e209131cea8a6fb |
| RPO | 79.1 | Not recorded | 68fa75fa1e209131cea6ac1e |
| zephyr-gemma-7b | 77.6 | Not recorded | 68fa75fa1e209131cea6ac1e |
| DPO | 75.7 | Not recorded | 68fa75fa1e209131cea6ac1e |
| zephyr-7b-gemma-sft | 75 | Not recorded | 68fa75fa1e209131cea6ac1e |

### LD-MoLE

ARC Easy Accuracy: 92.11. Result ID: 699e9d5e5698665ec4671b3f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bc81e209131cea8a6fb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7bc81e209131cea8a6fb.md>)

Recorded evaluation configuration: {"Model":"Qwen3-1.7B","TP":"4.23 %"}. This record does not assert independent reproduction.

### ReMoLE

ARC Easy Accuracy: 91.75. Result ID: 699e9d5e5698665ec4671b3e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bc81e209131cea8a6fb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa772c1e209131cea71959.md>)

Recorded evaluation configuration: {"Model":"Qwen3-1.7B","TP":"4.12 %"}. This record does not assert independent reproduction.

### ReMoLE

ARC Easy Accuracy: 91.61. Result ID: 699e9d5e5698665ec4671b36.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bc81e209131cea8a6fb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa772c1e209131cea71959.md>)

Recorded evaluation configuration: {"Model":"Llama3.1-8B","TP":"2.17 %"}. This record does not assert independent reproduction.

### LD-MoLE

ARC Easy Accuracy: 91.61. Result ID: 699e9d5e5698665ec4671b37.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bc81e209131cea8a6fb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7bc81e209131cea8a6fb.md>)

Recorded evaluation configuration: {"Model":"Llama3.1-8B","TP":"2.33 %"}. This record does not assert independent reproduction.

### LD-MoLE

ARC Easy Accuracy: 89.47. Result ID: 699e9d5e5698665ec4671b3b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bc81e209131cea8a6fb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7bc81e209131cea8a6fb.md>)

Recorded evaluation configuration: {"Model":"Llama3.2-3B","TP":"3.28 %"}. This record does not assert independent reproduction.

### ReMoLE

ARC Easy Accuracy: 89.3. Result ID: 699e9d5e5698665ec4671b3a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bc81e209131cea8a6fb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa772c1e209131cea71959.md>)

Recorded evaluation configuration: {"Model":"Llama3.2-3B","TP":"3.11 %"}. This record does not assert independent reproduction.

### MoLA(8888)

ARC Easy Accuracy: 88.6. Result ID: 699e9d5e5698665ec4671b3c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bc81e209131cea8a6fb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa723c1e209131cea55288.md>)

Recorded evaluation configuration: {"Model":"Qwen3-1.7B","TP":"4.12 %"}. This record does not assert independent reproduction.

### MoLA(2468)

ARC Easy Accuracy: 88.42. Result ID: 699e9d5e5698665ec4671b3d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bc81e209131cea8a6fb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa723c1e209131cea55288.md>)

Recorded evaluation configuration: {"Model":"Qwen3-1.7B","TP":"2.39 %"}. This record does not assert independent reproduction.

### MoLA(2468)

ARC Easy Accuracy: 87.37. Result ID: 699e9d5e5698665ec4671b35.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bc81e209131cea8a6fb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa723c1e209131cea55288.md>)

Recorded evaluation configuration: {"Model":"Llama3.1-8B","TP":"1.26 %"}. This record does not assert independent reproduction.

### MoLA(8888)

ARC Easy Accuracy: 86.32. Result ID: 699e9d5e5698665ec4671b34.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bc81e209131cea8a6fb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa723c1e209131cea55288.md>)

Recorded evaluation configuration: {"Model":"Llama3.1-8B","TP":"2.17 %"}. This record does not assert independent reproduction.

### MoLA(2468)

ARC Easy Accuracy: 83.86. Result ID: 699e9d5e5698665ec4671b39.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bc81e209131cea8a6fb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa723c1e209131cea55288.md>)

Recorded evaluation configuration: {"Model":"Llama3.2-3B","TP":"1.80 %"}. This record does not assert independent reproduction.

### MoLA(8888)

ARC Easy Accuracy: 83.51. Result ID: 699e9d5e5698665ec4671b38.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bc81e209131cea8a6fb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa723c1e209131cea55288.md>)

Recorded evaluation configuration: {"Model":"Llama3.2-3B","TP":"3.11 %"}. This record does not assert independent reproduction.

### RPO

ARC Easy Accuracy: 79.1. Result ID: 699408d13a6c3148491a8de5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75fa1e209131cea6ac1e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75fa1e209131cea6ac1e.md>)

Recorded evaluation configuration: {"Chain-of-thought":false,"Few-shots":false,"Backbone":"Gemma"}. This record does not assert independent reproduction.

### zephyr-gemma-7b

ARC Easy Accuracy: 77.6. Result ID: 699408d13a6c3148491a8de8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75fa1e209131cea6ac1e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71081e209131cea4ebc1.md>)

[Implementation](<https://github.com/huggingface/alignment-handbook>)

Recorded evaluation configuration: {"Chain-of-thought":false,"Few-shots":false}. This record does not assert independent reproduction.

### DPO

ARC Easy Accuracy: 75.7. Result ID: 699408d13a6c3148491a8de6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75fa1e209131cea6ac1e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74511e209131cea6105d.md>)

Recorded evaluation configuration: {"Chain-of-thought":false,"Few-shots":false,"Backbone":"Gemma"}. This record does not assert independent reproduction.

### zephyr-7b-gemma-sft

ARC Easy Accuracy: 75. Result ID: 699408d13a6c3148491a8de7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75fa1e209131cea6ac1e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72f11e209131cea59723.md>)

Recorded evaluation configuration: {"Chain-of-thought":false,"Few-shots":false,"Backbone":"Gemma"}. This record does not assert independent reproduction.

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