# Aggregated Downstream Evaluation on ANLI, BoolQ, Winogrande, RTE, PiQA, ARC-Easy, and ARC-Challenge

Source: [Aggregated Downstream Evaluation on ANLI, BoolQ, Winogrande, RTE, PiQA, ARC-Easy, and ARC-Challenge](<https://www.sota2.com/research/sota/aggregated-downstream-evaluation-on-anli-boolq-winogrande-rte-piqa-arc-easy-and-arc-challenge>)

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

Data updated: 2026-04-13T07:19:34.571Z

Evaluation across a suite of seven diverse natural language understanding and reasoning benchmarks: ANLI, BoolQ, Winogrande, RTE, PiQA, ARC-Easy, and ARC-Challenge.

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 | Aggregated Downstream Evaluation |
| Dataset (including any stated split) | ANLI, BoolQ, Winogrande, RTE, PiQA, ARC-Easy, ARC-Challenge |
| Metric | Average Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 8 of 8 |
| 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

- [Average Accuracy](<https://www.sota2.com/research/sota/aggregated-downstream-evaluation-on-anli-boolq-winogrande-rte-piqa-arc-easy-and-arc-challenge.md?metric=Average%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Average Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| CoA-LoRA | 61.94 | Not recorded | 68fa7bb01e209131cea89e6e |
| Shared-LoRA | 61.72 | Not recorded | 68fa7bb01e209131cea89e6e |
| QLoRA | 61.05 | Not recorded | 68fa7bb01e209131cea89e6e |
| LQ-LoRA | 60.81 | Not recorded | 68fa7bb01e209131cea89e6e |
| CoA-LoRA | 53.35 | Not recorded | 68fa7bb01e209131cea89e6e |
| LQ-LoRA | 51.9 | Not recorded | 68fa7bb01e209131cea89e6e |
| Shared-LoRA | 51.23 | Not recorded | 68fa7bb01e209131cea89e6e |
| QLoRA | 51.12 | Not recorded | 68fa7bb01e209131cea89e6e |

### CoA-LoRA

Average Accuracy: 61.94. Result ID: 69dc988d47f34ce924c32c2a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bb01e209131cea89e6e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7bb01e209131cea89e6e.md>)

Recorded evaluation configuration: {"Model":"Llama-2-7B","Average Bit Budget":"3-bit","Zero-shot":"true"}. This record does not assert independent reproduction.

### Shared-LoRA

Average Accuracy: 61.72. Result ID: 69dc988d47f34ce924c32c29.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bb01e209131cea89e6e.md>)

Recorded evaluation configuration: {"Model":"Llama-2-7B","Average Bit Budget":"3-bit","Zero-shot":"true"}. This record does not assert independent reproduction.

### QLoRA

Average Accuracy: 61.05. Result ID: 69dc988d47f34ce924c32c27.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bb01e209131cea89e6e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f6a1e209131cea456e1.md>)

[Implementation](<https://github.com/artidoro/qlora>)

Recorded evaluation configuration: {"Model":"Llama-2-7B","Average Bit Budget":"3-bit","Zero-shot":"true"}. This record does not assert independent reproduction.

### LQ-LoRA

Average Accuracy: 60.81. Result ID: 69dc988d47f34ce924c32c28.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bb01e209131cea89e6e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74a51e209131cea62e31.md>)

Recorded evaluation configuration: {"Model":"Llama-2-7B","Average Bit Budget":"3-bit","Zero-shot":"true"}. This record does not assert independent reproduction.

### CoA-LoRA

Average Accuracy: 53.35. Result ID: 69dc988d47f34ce924c32c26.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bb01e209131cea89e6e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7bb01e209131cea89e6e.md>)

Recorded evaluation configuration: {"Model":"Qwen2.5-3B","Average Bit Budget":"3-bit","Zero-shot":"true"}. This record does not assert independent reproduction.

### LQ-LoRA

Average Accuracy: 51.9. Result ID: 69dc988d47f34ce924c32c24.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bb01e209131cea89e6e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74a51e209131cea62e31.md>)

Recorded evaluation configuration: {"Model":"Qwen2.5-3B","Average Bit Budget":"3-bit","Zero-shot":"true"}. This record does not assert independent reproduction.

### Shared-LoRA

Average Accuracy: 51.23. Result ID: 69dc988d47f34ce924c32c25.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bb01e209131cea89e6e.md>)

Recorded evaluation configuration: {"Model":"Qwen2.5-3B","Average Bit Budget":"3-bit","Zero-shot":"true"}. This record does not assert independent reproduction.

### QLoRA

Average Accuracy: 51.12. Result ID: 69dc988d47f34ce924c32c23.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bb01e209131cea89e6e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f6a1e209131cea456e1.md>)

[Implementation](<https://github.com/artidoro/qlora>)

Recorded evaluation configuration: {"Model":"Qwen2.5-3B","Average Bit Budget":"3-bit","Zero-shot":"true"}. This record does not assert independent reproduction.

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