# Story Generation on HANNA

Source: [Story Generation on HANNA](<https://www.sota2.com/research/sota/story-generation-on-hanna>)

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

Data updated: 2026-03-17T07:09:55.629Z

Evaluation of story generation models on the HANNA dataset, assessed using Win Base, Win HANNA, and Average Score metrics.

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 | Story Generation |
| Dataset (including any stated split) | HANNA |
| Metric | Win HANNA Score |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 4 of 4 |
| 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

- [Win HANNA Score](<https://www.sota2.com/research/sota/story-generation-on-hanna.md?metric=Win%20HANNA%20Score>): Higher is better; unit: PERCENT.

- [Win Base Score](<https://www.sota2.com/research/sota/story-generation-on-hanna.md?metric=Win%20Base%20Score>): Higher is better; unit: PERCENT.

- [Average Score](<https://www.sota2.com/research/sota/story-generation-on-hanna.md?metric=Average%20Score>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Win HANNA Score | Result date | Report paper |
| --- | --- | --- | --- |
| EvolvR GRPO | 31.62 | Not recorded | 68fa7a871e209131cea83b76 |
| Qwen2.5-7B-Instruct + Point-RM GRPO | 29.09 | Not recorded | 68fa7a871e209131cea83b76 |
| Qwen2.5-7B-Instruct | 22.49 | Not recorded | 68fa7a871e209131cea83b76 |
| Qwen2.5-7B-Instruct + SFT | 17.67 | Not recorded | 68fa7a871e209131cea83b76 |

### EvolvR GRPO

Win HANNA Score: 31.62. Result ID: 69b8fd2c47f34ce924a0f58c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a871e209131cea83b76.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a871e209131cea83b76.md>)

Recorded evaluation configuration: {"training":"EvolvR GRPO"}. This record does not assert independent reproduction.

### Qwen2.5-7B-Instruct + Point-RM GRPO

Win HANNA Score: 29.09. Result ID: 69b8fd2c47f34ce924a0f58b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a871e209131cea83b76.md>)

Recorded evaluation configuration: {"training":"Point-RM GRPO"}. This record does not assert independent reproduction.

### Qwen2.5-7B-Instruct

Win HANNA Score: 22.49. Result ID: 69b8fd2c47f34ce924a0f589.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a871e209131cea83b76.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74bf1e209131cea6378d.md>)

[Implementation](<https://github.com/QwenLM/Qwen2>)

Recorded evaluation configuration: {"status":"untuned base model"}. This record does not assert independent reproduction.

### Qwen2.5-7B-Instruct + SFT

Win HANNA Score: 17.67. Result ID: 69b8fd2c47f34ce924a0f58a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a871e209131cea83b76.md>)

Recorded evaluation configuration: {"training":"SFT"}. This record does not assert independent reproduction.

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