SPB Git

Repository Public Research preview

summary.md 1.6 KB

system n chrf edit_sim rougeL exact_match number_recall placeholders_exact
B0_identity 266 80.318 [79.115, 81.710] 0.784 [0.772, 0.796] 0.803 [0.791, 0.817] 0.000 [0.000, 0.000] 1.000 [1.000, 1.000] 1.000 [1.000, 1.000]
B1_generic_paraphrase 266 75.817 [74.539, 77.263] 0.660 [0.643, 0.677] 0.708 [0.693, 0.725] 0.000 [0.000, 0.000] 0.998 [0.995, 1.000] 1.000 [1.000, 1.000]
htr-m1-small 266 77.788 [76.652, 79.146] 0.756 [0.744, 0.770] 0.771 [0.759, 0.786] 0.000 [0.000, 0.000] 0.992 [0.984, 0.998] 1.000 [1.000, 1.000]
htr-m1-tiny-s3 266 77.936 [76.862, 79.211] 0.759 [0.748, 0.771] 0.773 [0.763, 0.787] 0.000 [0.000, 0.000] 0.998 [0.996, 1.000] 0.992 [0.973, 1.000]
htr-m1-tiny 266 77.901 [76.861, 79.126] 0.759 [0.748, 0.770] 0.773 [0.762, 0.786] 0.000 [0.000, 0.000] 0.994 [0.987, 0.998] 0.989 [0.968, 1.000]
htr-m3tokenonly-tiny 266 78.208 [77.079, 79.485] 0.762 [0.751, 0.774] 0.778 [0.767, 0.792] 0.000 [0.000, 0.000] 0.990 [0.979, 0.998] 0.989 [0.974, 1.000]

Stylometric W1 (mean over features, standardised; lower = closer to originals):

  • rewrite vs original: 0.190
  • B0_identity vs original: 0.190 · probe P(original) mean 0.42
  • B1_generic_paraphrase vs original: 0.205 · probe P(original) mean 0.49
  • htr-m1-small vs original: 0.165 · probe P(original) mean 0.46
  • htr-m1-tiny-s3 vs original: 0.168 · probe P(original) mean 0.47
  • htr-m1-tiny vs original: 0.173 · probe P(original) mean 0.47
  • htr-m3tokenonly-tiny vs original: 0.171 · probe P(original) mean 0.46
Files main
Menu

Theme