# Natural Language Inference on FOLIO

Source: [Natural Language Inference on FOLIO](<https://www.sota2.com/research/sota/natural-language-inference-on-folio>)

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

Data updated: 2026-02-26T05:32:51.318Z

Evaluation of Natural Language Inference performance on the FOLIO dataset, measured by Accuracy.

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 | Natural Language Inference |
| Dataset (including any stated split) | FOLIO |
| Metric | Accuracy |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 26 of 26 |
| 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/natural-language-inference-on-folio.md?metric=Accuracy>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| gemma-2-27b | 0.61 | Not recorded | 6993f6239ebc499c97409b3a |
| gemma-2-9b | 0.53 | Not recorded | 6993f6239ebc499c97409b3a |
| gemma-2-27b | 0.4 | Not recorded | 6993f6239ebc499c97409b3a |
| gemma-2-9b | 0.38 | Not recorded | 6993f6239ebc499c97409b3a |
| llama-3.2-1b | 0.38 | Not recorded | 6993f6239ebc499c97409b3a |
| mistral-22b | 0.38 | Not recorded | 6993f6239ebc499c97409b3a |
| llama-3.1-8b | 0.37 | Not recorded | 6993f6239ebc499c97409b3a |
| ministral-8b | 0.37 | Not recorded | 6993f6239ebc499c97409b3a |
| qwen-2.5-7b | 0.37 | Not recorded | 6993f6239ebc499c97409b3a |
| qwen-2.5-14b | 0.37 | Not recorded | 6993f6239ebc499c97409b3a |
| llama-3.2-3b | 0.36 | Not recorded | 6993f6239ebc499c97409b3a |
| gemma-2-2b | 0.35 | Not recorded | 6993f6239ebc499c97409b3a |
| qwen-2.5-1.5b | 0.34 | Not recorded | 6993f6239ebc499c97409b3a |
| qwen-2.5-14b | 0.34 | Not recorded | 6993f6239ebc499c97409b3a |
| qwen-2.5-3b | 0.32 | Not recorded | 6993f6239ebc499c97409b3a |
| gemma-2-2b | 0.27 | Not recorded | 6993f6239ebc499c97409b3a |
| qwen-2.5-0.5b | 0.26 | Not recorded | 6993f6239ebc499c97409b3a |
| mistral-22b | 0.23 | Not recorded | 6993f6239ebc499c97409b3a |
| qwen-2.5-1.5b | 0.19 | Not recorded | 6993f6239ebc499c97409b3a |
| llama-3.2-3b | 0.18 | Not recorded | 6993f6239ebc499c97409b3a |
| llama-3.2-1b | 0.17 | Not recorded | 6993f6239ebc499c97409b3a |
| qwen-2.5-0.5b | 0.17 | Not recorded | 6993f6239ebc499c97409b3a |
| ministral-8b | 0.16 | Not recorded | 6993f6239ebc499c97409b3a |
| llama-3.1-8b | 0.12 | Not recorded | 6993f6239ebc499c97409b3a |
| qwen-2.5-3b | 0.08 | Not recorded | 6993f6239ebc499c97409b3a |
| qwen-2.5-7b | 0.01 | Not recorded | 6993f6239ebc499c97409b3a |

### gemma-2-27b

Accuracy: 0.61. Result ID: 699a70b15698665ec4f9772f.

[Reporting paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

Recorded evaluation configuration: {"Decoding Method":"GCD"}. This record does not assert independent reproduction.

### gemma-2-9b

Accuracy: 0.53. Result ID: 699a70b15698665ec4f9772d.

[Reporting paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

Recorded evaluation configuration: {"Decoding Method":"GCD"}. This record does not assert independent reproduction.

### gemma-2-27b

Accuracy: 0.4. Result ID: 699a70b15698665ec4f97730.

[Reporting paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

Recorded evaluation configuration: {"Decoding Method":"NL2LOGIC"}. This record does not assert independent reproduction.

### gemma-2-9b

Accuracy: 0.38. Result ID: 699a70b15698665ec4f9772e.

[Reporting paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

Recorded evaluation configuration: {"Decoding Method":"NL2LOGIC"}. This record does not assert independent reproduction.

### llama-3.2-1b

Accuracy: 0.38. Result ID: 699a70b15698665ec4f97732.

[Reporting paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

Recorded evaluation configuration: {"Decoding Method":"NL2LOGIC"}. This record does not assert independent reproduction.

### mistral-22b

Accuracy: 0.38. Result ID: 699a70b15698665ec4f9773a.

[Reporting paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

Recorded evaluation configuration: {"Decoding Method":"NL2LOGIC"}. This record does not assert independent reproduction.

### llama-3.1-8b

Accuracy: 0.37. Result ID: 699a70b15698665ec4f97736.

[Reporting paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

Recorded evaluation configuration: {"Decoding Method":"NL2LOGIC"}. This record does not assert independent reproduction.

### ministral-8b

Accuracy: 0.37. Result ID: 699a70b15698665ec4f97738.

[Reporting paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

Recorded evaluation configuration: {"Decoding Method":"NL2LOGIC"}. This record does not assert independent reproduction.

### qwen-2.5-7b

Accuracy: 0.37. Result ID: 699a70b15698665ec4f97742.

[Reporting paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

Recorded evaluation configuration: {"Decoding Method":"NL2LOGIC"}. This record does not assert independent reproduction.

### qwen-2.5-14b

Accuracy: 0.37. Result ID: 699a70b15698665ec4f97744.

[Reporting paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

Recorded evaluation configuration: {"Decoding Method":"NL2LOGIC"}. This record does not assert independent reproduction.

### llama-3.2-3b

Accuracy: 0.36. Result ID: 699a70b15698665ec4f97734.

[Reporting paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

Recorded evaluation configuration: {"Decoding Method":"NL2LOGIC"}. This record does not assert independent reproduction.

### gemma-2-2b

Accuracy: 0.35. Result ID: 699a70b15698665ec4f9772c.

[Reporting paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

Recorded evaluation configuration: {"Decoding Method":"NL2LOGIC"}. This record does not assert independent reproduction.

### qwen-2.5-1.5b

Accuracy: 0.34. Result ID: 699a70b15698665ec4f9773e.

[Reporting paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

Recorded evaluation configuration: {"Decoding Method":"NL2LOGIC"}. This record does not assert independent reproduction.

### qwen-2.5-14b

Accuracy: 0.34. Result ID: 699a70b15698665ec4f97743.

[Reporting paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

Recorded evaluation configuration: {"Decoding Method":"GCD"}. This record does not assert independent reproduction.

### qwen-2.5-3b

Accuracy: 0.32. Result ID: 699a70b15698665ec4f97740.

[Reporting paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

Recorded evaluation configuration: {"Decoding Method":"NL2LOGIC"}. This record does not assert independent reproduction.

### gemma-2-2b

Accuracy: 0.27. Result ID: 699a70b15698665ec4f9772b.

[Reporting paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

Recorded evaluation configuration: {"Decoding Method":"GCD"}. This record does not assert independent reproduction.

### qwen-2.5-0.5b

Accuracy: 0.26. Result ID: 699a70b15698665ec4f9773c.

[Reporting paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

Recorded evaluation configuration: {"Decoding Method":"NL2LOGIC"}. This record does not assert independent reproduction.

### mistral-22b

Accuracy: 0.23. Result ID: 699a70b15698665ec4f97739.

[Reporting paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

Recorded evaluation configuration: {"Decoding Method":"GCD"}. This record does not assert independent reproduction.

### qwen-2.5-1.5b

Accuracy: 0.19. Result ID: 699a70b15698665ec4f9773d.

[Reporting paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

Recorded evaluation configuration: {"Decoding Method":"GCD"}. This record does not assert independent reproduction.

### llama-3.2-3b

Accuracy: 0.18. Result ID: 699a70b15698665ec4f97733.

[Reporting paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

Recorded evaluation configuration: {"Decoding Method":"GCD"}. This record does not assert independent reproduction.

### llama-3.2-1b

Accuracy: 0.17. Result ID: 699a70b15698665ec4f97731.

[Reporting paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

Recorded evaluation configuration: {"Decoding Method":"GCD"}. This record does not assert independent reproduction.

### qwen-2.5-0.5b

Accuracy: 0.17. Result ID: 699a70b15698665ec4f9773b.

[Reporting paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

Recorded evaluation configuration: {"Decoding Method":"GCD"}. This record does not assert independent reproduction.

### ministral-8b

Accuracy: 0.16. Result ID: 699a70b15698665ec4f97737.

[Reporting paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

Recorded evaluation configuration: {"Decoding Method":"GCD"}. This record does not assert independent reproduction.

### llama-3.1-8b

Accuracy: 0.12. Result ID: 699a70b15698665ec4f97735.

[Reporting paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

Recorded evaluation configuration: {"Decoding Method":"GCD"}. This record does not assert independent reproduction.

### qwen-2.5-3b

Accuracy: 0.08. Result ID: 699a70b15698665ec4f9773f.

[Reporting paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

Recorded evaluation configuration: {"Decoding Method":"GCD"}. This record does not assert independent reproduction.

### qwen-2.5-7b

Accuracy: 0.01. Result ID: 699a70b15698665ec4f97741.

[Reporting paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6993f6239ebc499c97409b3a.md>)

Recorded evaluation configuration: {"Decoding Method":"GCD"}. This record does not assert independent reproduction.

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