### gemma2-9b · full · geo
   idiom config:    {'model': 'google/gemma-2-9b', 'reduction': 'geometric_mean', 'medial_only': False, 'dtype': 'bfloat16', 'dataset': '/home/prada/PID_evaluation/data/dataset.tsv', 'num_idioms': 18, 'syn_reg_eps': 0.01}
   nonidiom config: {'model': 'google/gemma-2-9b', 'reduction': 'geometric_mean', 'medial_only': False, '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
   1.1968     1.1956  [  1.1676,  1.2257]

== non-idioms :: ratio_u_idiom ==  (N=18 phrases)
     mean     median                    95% CI
   1.0748     1.0674  [  1.0564,  1.0944]
  cross-dataset ratio_u_idiom: idioms - nonidioms  Δ=+0.1220  CI=[+0.0871,+0.1568]  *

== idioms :: ratio_s_idiom ==  (N=9 phrases)  (9 non-finite dropped)
     mean     median                    95% CI
   1.1480     1.1537  [  1.1218,  1.1733]

== non-idioms :: ratio_s_idiom ==  (N=1 phrases)  (17 non-finite dropped)
     mean     median                    95% CI
   1.2136     1.2136  [  1.2136,  1.2136]
  cross-dataset ratio_s_idiom: idioms - nonidioms  Δ=-0.0657  CI=[-0.0919,-0.0405]  *