model: gpt-5.4-2026-03-05
judge_model: gpt-4o-mini-2024-07-18
inference_model: gpt-4o-mini-2024-07-18
mode: few_shot
strategy: full
n_hypotheses: 10
n_samples: 60
n_test_instances: 100
n_tasks: 7
avg_clarity: 3.586 +/- 0.212
avg_novelty: 3.457 +/- 0.282
avg_plausibility: 3.942 +/- 0.14
avg_quality: 3.662 +/- 0.153
avg_diversity: 0.334 +/- 0.049… See the full description on the dataset page:
https://huggingface.co/datasets/strategy-scope/hypobench-few_shot-gpt5420260305-full-20260415_015532.
