# Generative Commonsense Reasoning on Dense-ATOMIC 500 manually selected paths (test)

Source: [Generative Commonsense Reasoning on Dense-ATOMIC 500 manually selected paths (test)](<https://www.sota2.com/research/sota/generative-commonsense-reasoning-on-dense-atomic-500-manually-selected-paths-test>)

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

Data updated: 2026-02-26T05:37:12.609Z

Evaluation of model performance on generative commonsense reasoning using the test split of the Dense-ATOMIC 500 dataset, measured by BLEU and ROUGE scores.

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 | Generative Commonsense Reasoning |
| Dataset (including any stated split) | Dense-ATOMIC 500 manually selected paths (test) |
| Metric | BLEU-1 |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 2 of 2 |
| 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

- [BLEU-1](<https://www.sota2.com/research/sota/generative-commonsense-reasoning-on-dense-atomic-500-manually-selected-paths-test.md?metric=BLEU-1>): Higher is better; unit: PERCENT.

- [BLEU-2](<https://www.sota2.com/research/sota/generative-commonsense-reasoning-on-dense-atomic-500-manually-selected-paths-test.md?metric=BLEU-2>): Higher is better; unit: PERCENT.

- [ROUGE-L](<https://www.sota2.com/research/sota/generative-commonsense-reasoning-on-dense-atomic-500-manually-selected-paths-test.md?metric=ROUGE-L>): Higher is better; unit: PERCENT.

## Recorded results

| Method | BLEU-1 | Result date | Report paper |
| --- | --- | --- | --- |
| BART | 48.63 | Not recorded | 68fa6f761e209131cea45bc2 |
| BART | 48.57 | Not recorded | 68fa6f761e209131cea45bc2 |

### BART

BLEU-1: 48.63. Result ID: 6990c10e3a6c31484903033f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6f761e209131cea45bc2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66921e209131cea10fff.md>)

Recorded evaluation configuration: {"Input settings":"Multi-hop","Input details":"given the first four nodes of the path"}. This record does not assert independent reproduction.

### BART

BLEU-1: 48.57. Result ID: 6990c10e3a6c31484903033e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6f761e209131cea45bc2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66921e209131cea10fff.md>)

Recorded evaluation configuration: {"Input settings":"One-hop","Input details":"given only the first node of the path"}. This record does not assert independent reproduction.

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