### qwen3-8b-base · medial · geo
   idiom config:    {'model': 'Qwen/Qwen3-8B-Base', 'reduction': 'geometric_mean', 'medial_only': True, 'dtype': 'bfloat16', 'dataset': '/home/prada/PID_evaluation/data/dataset.tsv', 'num_idioms': 18, 'syn_reg_eps': 0.01}
   nonidiom config: {'model': 'Qwen/Qwen3-8B-Base', 'reduction': 'geometric_mean', 'medial_only': True, 'dtype': 'bfloat16', 'dataset': '/home/prada/PID_evaluation/data/nonidioms_dataset.tsv', 'num_idioms': 18, 'syn_reg_eps': 0.01}

== idioms :: ratio_u_idiom ==  (N=18 phrases)
     mean     median                    95% CI
   1.2173     1.2044  [  1.1808,  1.2555]

== non-idioms :: ratio_u_idiom ==  (N=18 phrases)
     mean     median                    95% CI
   1.0764     1.0799  [  1.0598,  1.0935]
  cross-dataset ratio_u_idiom: idioms - nonidioms  Δ=+0.1409  CI=[+0.1000,+0.1824]  *

== idioms :: ratio_s_idiom ==  (N=15 phrases)  (3 non-finite dropped)
     mean     median                    95% CI
   1.1773     1.1641  [  1.1437,  1.2131]

== non-idioms :: ratio_s_idiom ==  (N=4 phrases)  (14 non-finite dropped)
     mean     median                    95% CI
   1.3542     1.3076  [  1.2456,  1.5093]
  cross-dataset ratio_s_idiom: idioms - nonidioms  Δ=-0.1769  CI=[-0.3293,-0.0618]  *