Model outputs from the micro-creativity inference suite.
Model: Qwen/Qwen3-8B
Dataset: arena-hard-creative-writing (250 items)
Part of collection: ZachW/llm-creativity-benchmarks
input
The exact prompt sent to the model (after meta-prompt application)… See the full description on the dataset page:
https://huggingface.co/datasets/ZachW/qwen3-8b_arena-hard-creative-writing.