model: gpt-5-mini-2025-08-07
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: 3.78 +/- 0.188
avg_novelty: 3.395 +/- 0.35
avg_plausibility: 4.104 +/- 0.181
avg_quality: 3.76 +/- 0.114
avg_diversity: 0.492 +/- 0.119… See the full description on the dataset page:
https://huggingface.co/datasets/strategy-scope/hypobench-few_shot-gpt5mini20250807-full-20260414_175807.
