# Logical Reasoning on Logical Reasoning Datasets Balanced

Source: [Logical Reasoning on Logical Reasoning Datasets Balanced](<https://www.sota2.com/research/sota/logical-reasoning-on-logical-reasoning-datasets-balanced>)

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

Data updated: 2026-02-26T06:01:45.114Z

Evaluation of model performance on the Logical Reasoning task using the balanced subset of Logical Reasoning Datasets, reporting scores across various sub-metrics like Recolr, LogiQA, and LogicBench.

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 | Logical Reasoning |
| Dataset (including any stated split) | Logical Reasoning Datasets Balanced |
| Metric | Recolr |
| 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

- [Recolr](<https://www.sota2.com/research/sota/logical-reasoning-on-logical-reasoning-datasets-balanced.md?metric=Recolr>): Higher is better; unit: PERCENT.

- [LogiQA](<https://www.sota2.com/research/sota/logical-reasoning-on-logical-reasoning-datasets-balanced.md?metric=LogiQA>): Higher is better; unit: PERCENT.

- [LogiQA 2.0](<https://www.sota2.com/research/sota/logical-reasoning-on-logical-reasoning-datasets-balanced.md?metric=LogiQA%202.0>): Higher is better; unit: PERCENT.

- [LogicBench](<https://www.sota2.com/research/sota/logical-reasoning-on-logical-reasoning-datasets-balanced.md?metric=LogicBench>): Higher is better; unit: PERCENT.

- [Average Score](<https://www.sota2.com/research/sota/logical-reasoning-on-logical-reasoning-datasets-balanced.md?metric=Average%20Score>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Recolr | Result date | Report paper |
| --- | --- | --- | --- |
| Our Re-Cognizing Method | 94.8 | Not recorded | 68fa7be71e209131cea8b361 |
| Our Re-Cognizing Method | 92.2 | Not recorded | 68fa7be71e209131cea8b361 |
| Optimization w/o Stage1 | 91.8 | Not recorded | 68fa7be71e209131cea8b361 |
| Optimization w/o Stage2 | 91.8 | Not recorded | 68fa7be71e209131cea8b361 |
| Bin-based Progressive Learning | 91.1 | Not recorded | 68fa7be71e209131cea8b361 |
| Curriculum Learning | 90.8 | Not recorded | 68fa7be71e209131cea8b361 |
| Directly | 90.1 | Not recorded | 68fa7be71e209131cea8b361 |
| Bin-based Progressive Learning | 89.4 | Not recorded | 68fa7be71e209131cea8b361 |
| Curriculum Learning | 87.2 | Not recorded | 68fa7be71e209131cea8b361 |
| Optimization w/o Stage2 | 83.9 | Not recorded | 68fa7be71e209131cea8b361 |
| Optimization w/o Stage1 | 83.3 | Not recorded | 68fa7be71e209131cea8b361 |
| Directly | 79.7 | Not recorded | 68fa7be71e209131cea8b361 |
| GPT-4 | 79.4 | Not recorded | 68fa7be71e209131cea8b361 |
| DeepSeek-V3 | 73.8 | Not recorded | 68fa7be71e209131cea8b361 |
| Base | 44.7 | Not recorded | 68fa7be71e209131cea8b361 |
| Base | 43.3 | Not recorded | 68fa7be71e209131cea8b361 |

### Our Re-Cognizing Method

Recolr: 94.8. Result ID: 699a3f435698665ec4c5f473.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b361.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b361.md>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B"}. This record does not assert independent reproduction.

### Our Re-Cognizing Method

Recolr: 92.2. Result ID: 699a3f435698665ec4c5f46c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b361.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b361.md>)

Recorded evaluation configuration: {"Backbone":"LLaMA3.1-8B"}. This record does not assert independent reproduction.

### Optimization w/o Stage1

Recolr: 91.8. Result ID: 699a3f435698665ec4c5f46a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b361.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b361.md>)

Recorded evaluation configuration: {"Backbone":"LLaMA3.1-8B"}. This record does not assert independent reproduction.

### Optimization w/o Stage2

Recolr: 91.8. Result ID: 699a3f435698665ec4c5f46b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b361.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b361.md>)

Recorded evaluation configuration: {"Backbone":"LLaMA3.1-8B"}. This record does not assert independent reproduction.

### Bin-based Progressive Learning

Recolr: 91.1. Result ID: 699a3f435698665ec4c5f470.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b361.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa5c081e209131ce9cda9b.md>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B"}. This record does not assert independent reproduction.

### Curriculum Learning

Recolr: 90.8. Result ID: 699a3f435698665ec4c5f468.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b361.md>)

Recorded evaluation configuration: {"Backbone":"LLaMA3.1-8B"}. This record does not assert independent reproduction.

### Directly

Recolr: 90.1. Result ID: 699a3f435698665ec4c5f467.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b361.md>)

Recorded evaluation configuration: {"Backbone":"LLaMA3.1-8B"}. This record does not assert independent reproduction.

### Bin-based Progressive Learning

Recolr: 89.4. Result ID: 699a3f435698665ec4c5f469.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b361.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa5c081e209131ce9cda9b.md>)

Recorded evaluation configuration: {"Backbone":"LLaMA3.1-8B"}. This record does not assert independent reproduction.

### Curriculum Learning

Recolr: 87.2. Result ID: 699a3f435698665ec4c5f46f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b361.md>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B"}. This record does not assert independent reproduction.

### Optimization w/o Stage2

Recolr: 83.9. Result ID: 699a3f435698665ec4c5f472.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b361.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b361.md>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B"}. This record does not assert independent reproduction.

### Optimization w/o Stage1

Recolr: 83.3. Result ID: 699a3f435698665ec4c5f471.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b361.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b361.md>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B"}. This record does not assert independent reproduction.

### Directly

Recolr: 79.7. Result ID: 699a3f435698665ec4c5f46e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b361.md>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B"}. This record does not assert independent reproduction.

### GPT-4

Recolr: 79.4. Result ID: 699a3f435698665ec4c5f464.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b361.md>)

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

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

Recorded evaluation configuration: {"Model Category":"Closed-source LLMs"}. This record does not assert independent reproduction.

### DeepSeek-V3

Recolr: 73.8. Result ID: 699a3f435698665ec4c5f465.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b361.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77001e209131cea7066a.md>)

Recorded evaluation configuration: {"Model Category":"Closed-source LLMs"}. This record does not assert independent reproduction.

### Base

Recolr: 44.7. Result ID: 699a3f435698665ec4c5f466.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b361.md>)

Recorded evaluation configuration: {"Backbone":"LLaMA3.1-8B"}. This record does not assert independent reproduction.

### Base

Recolr: 43.3. Result ID: 699a3f435698665ec4c5f46d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7be71e209131cea8b361.md>)

Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B"}. This record does not assert independent reproduction.

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