# Syllogistic Reasoning on Syllogistic Reasoning (test)

Source: [Syllogistic Reasoning on Syllogistic Reasoning (test)](<https://www.sota2.com/research/sota/syllogistic-reasoning-on-syllogistic-reasoning-test>)

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

Data updated: 2026-03-09T06:49:15.502Z

Evaluation of model performance on Syllogistic Reasoning using the test split of the Syllogistic Reasoning dataset, measured by Accuracy (Acc) and Confidence Error (CE).

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 | Syllogistic Reasoning |
| Dataset (including any stated split) | Syllogistic Reasoning (test) |
| Metric | Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 32 of 32 |
| 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/syllogistic-reasoning-on-syllogistic-reasoning-test.md?metric=Accuracy>): Higher is better; unit: PERCENT.

- [Confidence Error](<https://www.sota2.com/research/sota/syllogistic-reasoning-on-syllogistic-reasoning-test.md?metric=Confidence%20Error>): Lower is better; unit: PERCENT.

- [Accuracy / Confidence Error Ratio](<https://www.sota2.com/research/sota/syllogistic-reasoning-on-syllogistic-reasoning-test.md?metric=Accuracy%20%2F%20Confidence%20Error%20Ratio>): Higher is better; unit: SCALAR.

- [Absolute Delta (Acc vs CE)](<https://www.sota2.com/research/sota/syllogistic-reasoning-on-syllogistic-reasoning-test.md?metric=Absolute%20Delta%20(Acc%20vs%20CE)>): Lower is better; unit: SCALAR.

- [Relative Delta (Acc vs CE)](<https://www.sota2.com/research/sota/syllogistic-reasoning-on-syllogistic-reasoning-test.md?metric=Relative%20Delta%20(Acc%20vs%20CE)>): Lower is better; unit: SCALAR.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| Contrastive Steering (Qwen 2.5) | 89.9 | Not recorded | 68fa78991e209131cea7939d |
| Contrastive Steering (Qwen 2.5) | 89.9 | Not recorded | 68fa78991e209131cea7939d |
| Qwen 2.5 | 89.8 | Not recorded | 68fa78991e209131cea7939d |
| Qwen 2.5 | 88.85 | Not recorded | 68fa78991e209131cea7939d |
| Contrastive Steering (Qwen 2.5) | 86.63 | Not recorded | 68fa78991e209131cea7939d |
| Qwen 2.5 | 86.44 | Not recorded | 68fa78991e209131cea7939d |
| Qwen 2.5 | 85.29 | Not recorded | 68fa78991e209131cea7939d |
| Contrastive Steering (Qwen 2.5) | 85.29 | Not recorded | 68fa78991e209131cea7939d |
| Contrastive Steering (Llama 3.1) | 85.1 | Not recorded | 68fa78991e209131cea7939d |
| Gemma 2 | 85 | Not recorded | 68fa78991e209131cea7939d |
| Gemma 2 | 84.61 | Not recorded | 68fa78991e209131cea7939d |
| Contrastive Steering (Gemma 2) | 83.27 | Not recorded | 68fa78991e209131cea7939d |
| Contrastive Steering (Gemma 2) | 80.38 | Not recorded | 68fa78991e209131cea7939d |
| Llama 3.1 | 78.27 | Not recorded | 68fa78991e209131cea7939d |
| Contrastive Steering (Llama 3.2) | 78.27 | Not recorded | 68fa78991e209131cea7939d |
| Llama 3.2 | 77.79 | Not recorded | 68fa78991e209131cea7939d |
| Contrastive Steering (Llama 3.2) | 77.79 | Not recorded | 68fa78991e209131cea7939d |
| Contrastive Steering (Qwen 2.5) | 77.79 | Not recorded | 68fa78991e209131cea7939d |
| Qwen 2.5 | 75.67 | Not recorded | 68fa78991e209131cea7939d |
| Contrastive Steering (Gemma 2) | 74.13 | Not recorded | 68fa78991e209131cea7939d |
| Contrastive Steering (Llama 3.2) | 73.56 | Not recorded | 68fa78991e209131cea7939d |
| Gemma 2 | 73.27 | Not recorded | 68fa78991e209131cea7939d |
| Llama 3.2 | 72.79 | Not recorded | 68fa78991e209131cea7939d |
| Contrastive Steering (Qwen 2.5) | 72.69 | Not recorded | 68fa78991e209131cea7939d |
| Contrastive Steering (Gemma 2) | 70 | Not recorded | 68fa78991e209131cea7939d |
| Gemma 2 | 69.42 | Not recorded | 68fa78991e209131cea7939d |
| Contrastive Steering (Llama 3.2) | 66.44 | Not recorded | 68fa78991e209131cea7939d |
| Llama 3.2 | 58.17 | Not recorded | 68fa78991e209131cea7939d |
| Llama 3.2 | 57.21 | Not recorded | 68fa78991e209131cea7939d |
| Qwen 2.5 | 51.92 | Not recorded | 68fa78991e209131cea7939d |
| Llama 3.1 | 40.58 | Not recorded | 68fa78991e209131cea7939d |
| Contrastive Steering (Llama 3.1) | 30.67 | Not recorded | 68fa78991e209131cea7939d |

### Contrastive Steering (Qwen 2.5)

Accuracy: 89.9. Result ID: 69ae6b606dbb7075b955df9e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

Recorded evaluation configuration: {"Size":"7b","Setting":"Zero-shot","Steering":"Steered Model (αbest)","αbest":"-1.5"}. This record does not assert independent reproduction.

### Contrastive Steering (Qwen 2.5)

Accuracy: 89.9. Result ID: 69ae6b606dbb7075b955dfae.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

Recorded evaluation configuration: {"Size":"7b","Setting":"Few-shot","Steering":"Steered Model (αbest)","αbest":"0.9"}. This record does not assert independent reproduction.

### Qwen 2.5

Accuracy: 89.8. Result ID: 69ae6b606dbb7075b955dfad.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70b91e209131cea4cdc5.md>)

Recorded evaluation configuration: {"Size":"7b","Setting":"Few-shot","Steering":"Base Model (α = 0)"}. This record does not assert independent reproduction.

### Qwen 2.5

Accuracy: 88.85. Result ID: 69ae6b606dbb7075b955df9d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70b91e209131cea4cdc5.md>)

Recorded evaluation configuration: {"Size":"7b","Setting":"Zero-shot","Steering":"Base Model (α = 0)"}. This record does not assert independent reproduction.

### Contrastive Steering (Qwen 2.5)

Accuracy: 86.63. Result ID: 69ae6b606dbb7075b955dfac.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

Recorded evaluation configuration: {"Size":"3b","Setting":"Few-shot","Steering":"Steered Model (αbest)","αbest":"0.6"}. This record does not assert independent reproduction.

### Qwen 2.5

Accuracy: 86.44. Result ID: 69ae6b606dbb7075b955dfab.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70b91e209131cea4cdc5.md>)

Recorded evaluation configuration: {"Size":"3b","Setting":"Few-shot","Steering":"Base Model (α = 0)"}. This record does not assert independent reproduction.

### Qwen 2.5

Accuracy: 85.29. Result ID: 69ae6b606dbb7075b955df9b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70b91e209131cea4cdc5.md>)

Recorded evaluation configuration: {"Size":"3b","Setting":"Zero-shot","Steering":"Base Model (α = 0)"}. This record does not assert independent reproduction.

### Contrastive Steering (Qwen 2.5)

Accuracy: 85.29. Result ID: 69ae6b606dbb7075b955df9c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

Recorded evaluation configuration: {"Size":"3b","Setting":"Zero-shot","Steering":"Steered Model (αbest)","αbest":"0.0"}. This record does not assert independent reproduction.

### Contrastive Steering (Llama 3.1)

Accuracy: 85.1. Result ID: 69ae6b606dbb7075b955df94.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

Recorded evaluation configuration: {"Size":"8b","Setting":"Zero-shot","Steering":"Steered Model (αbest)","αbest":"0.9"}. This record does not assert independent reproduction.

### Gemma 2

Accuracy: 85. Result ID: 69ae6b606dbb7075b955df97.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75001e209131cea6527c.md>)

Recorded evaluation configuration: {"Size":"9b","Setting":"Zero-shot","Steering":"Base Model (α = 0)"}. This record does not assert independent reproduction.

### Gemma 2

Accuracy: 84.61. Result ID: 69ae6b606dbb7075b955dfa7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75001e209131cea6527c.md>)

Recorded evaluation configuration: {"Size":"9b","Setting":"Few-shot","Steering":"Base Model (α = 0)"}. This record does not assert independent reproduction.

### Contrastive Steering (Gemma 2)

Accuracy: 83.27. Result ID: 69ae6b606dbb7075b955df98.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

Recorded evaluation configuration: {"Size":"9b","Setting":"Zero-shot","Steering":"Steered Model (αbest)","αbest":"0.6"}. This record does not assert independent reproduction.

### Contrastive Steering (Gemma 2)

Accuracy: 80.38. Result ID: 69ae6b606dbb7075b955dfa8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

Recorded evaluation configuration: {"Size":"9b","Setting":"Few-shot","Steering":"Steered Model (αbest)","αbest":"0.3"}. This record does not assert independent reproduction.

### Llama 3.1

Accuracy: 78.27. Result ID: 69ae6b606dbb7075b955df93.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

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

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

Recorded evaluation configuration: {"Size":"8b","Setting":"Zero-shot","Steering":"Base Model (α = 0)"}. This record does not assert independent reproduction.

### Contrastive Steering (Llama 3.2)

Accuracy: 78.27. Result ID: 69ae6b606dbb7075b955dfa2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

Recorded evaluation configuration: {"Size":"3b","Setting":"Few-shot","Steering":"Steered Model (αbest)","αbest":"0.3"}. This record does not assert independent reproduction.

### Llama 3.2

Accuracy: 77.79. Result ID: 69ae6b606dbb7075b955df91.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

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

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

Recorded evaluation configuration: {"Size":"3b","Setting":"Zero-shot","Steering":"Base Model (α = 0)"}. This record does not assert independent reproduction.

### Contrastive Steering (Llama 3.2)

Accuracy: 77.79. Result ID: 69ae6b606dbb7075b955df92.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

Recorded evaluation configuration: {"Size":"3b","Setting":"Zero-shot","Steering":"Steered Model (αbest)","αbest":"0.0"}. This record does not assert independent reproduction.

### Contrastive Steering (Qwen 2.5)

Accuracy: 77.79. Result ID: 69ae6b606dbb7075b955df9a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

Recorded evaluation configuration: {"Size":"1.5b","Setting":"Zero-shot","Steering":"Steered Model (αbest)","αbest":"0.3"}. This record does not assert independent reproduction.

### Qwen 2.5

Accuracy: 75.67. Result ID: 69ae6b606dbb7075b955df99.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70b91e209131cea4cdc5.md>)

Recorded evaluation configuration: {"Size":"1.5b","Setting":"Zero-shot","Steering":"Base Model (α = 0)"}. This record does not assert independent reproduction.

### Contrastive Steering (Gemma 2)

Accuracy: 74.13. Result ID: 69ae6b606dbb7075b955df96.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

Recorded evaluation configuration: {"Size":"2b","Setting":"Zero-shot","Steering":"Steered Model (αbest)","αbest":"1.8"}. This record does not assert independent reproduction.

### Contrastive Steering (Llama 3.2)

Accuracy: 73.56. Result ID: 69ae6b606dbb7075b955df90.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

Recorded evaluation configuration: {"Size":"1b","Setting":"Zero-shot","Steering":"Steered Model (αbest)","αbest":"-0.9"}. This record does not assert independent reproduction.

### Gemma 2

Accuracy: 73.27. Result ID: 69ae6b606dbb7075b955df95.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75001e209131cea6527c.md>)

Recorded evaluation configuration: {"Size":"2b","Setting":"Zero-shot","Steering":"Base Model (α = 0)"}. This record does not assert independent reproduction.

### Llama 3.2

Accuracy: 72.79. Result ID: 69ae6b606dbb7075b955dfa1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

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

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

Recorded evaluation configuration: {"Size":"3b","Setting":"Few-shot","Steering":"Base Model (α = 0)"}. This record does not assert independent reproduction.

### Contrastive Steering (Qwen 2.5)

Accuracy: 72.69. Result ID: 69ae6b606dbb7075b955dfaa.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

Recorded evaluation configuration: {"Size":"1.5b","Setting":"Few-shot","Steering":"Steered Model (αbest)","αbest":"-1.2"}. This record does not assert independent reproduction.

### Contrastive Steering (Gemma 2)

Accuracy: 70. Result ID: 69ae6b606dbb7075b955dfa6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

Recorded evaluation configuration: {"Size":"2b","Setting":"Few-shot","Steering":"Steered Model (αbest)","αbest":"-0.3"}. This record does not assert independent reproduction.

### Gemma 2

Accuracy: 69.42. Result ID: 69ae6b606dbb7075b955dfa5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75001e209131cea6527c.md>)

Recorded evaluation configuration: {"Size":"2b","Setting":"Few-shot","Steering":"Base Model (α = 0)"}. This record does not assert independent reproduction.

### Contrastive Steering (Llama 3.2)

Accuracy: 66.44. Result ID: 69ae6b606dbb7075b955dfa0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

Recorded evaluation configuration: {"Size":"1b","Setting":"Few-shot","Steering":"Steered Model (αbest)","αbest":"-1.5"}. This record does not assert independent reproduction.

### Llama 3.2

Accuracy: 58.17. Result ID: 69ae6b606dbb7075b955df8f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

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

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

Recorded evaluation configuration: {"Size":"1b","Setting":"Zero-shot","Steering":"Base Model (α = 0)"}. This record does not assert independent reproduction.

### Llama 3.2

Accuracy: 57.21. Result ID: 69ae6b606dbb7075b955df9f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

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

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

Recorded evaluation configuration: {"Size":"1b","Setting":"Few-shot","Steering":"Base Model (α = 0)"}. This record does not assert independent reproduction.

### Qwen 2.5

Accuracy: 51.92. Result ID: 69ae6b606dbb7075b955dfa9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70b91e209131cea4cdc5.md>)

Recorded evaluation configuration: {"Size":"1.5b","Setting":"Few-shot","Steering":"Base Model (α = 0)"}. This record does not assert independent reproduction.

### Llama 3.1

Accuracy: 40.58. Result ID: 69ae6b606dbb7075b955dfa3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

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

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

Recorded evaluation configuration: {"Size":"8b","Setting":"Few-shot","Steering":"Base Model (α = 0)"}. This record does not assert independent reproduction.

### Contrastive Steering (Llama 3.1)

Accuracy: 30.67. Result ID: 69ae6b606dbb7075b955dfa4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78991e209131cea7939d.md>)

Recorded evaluation configuration: {"Size":"8b","Setting":"Few-shot","Steering":"Steered Model (αbest)","αbest":"0.3"}. This record does not assert independent reproduction.

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