INDEX
Explanations
descriptive adjectives following 'to be' verbs
New Auto-Interp
Negative Logits
bastante
0.92
весьма
0.88
optimized
0.86
conducive
0.84
oldukça
0.83
颇
0.83
somewhat
0.83
꽤
0.83
quite
0.83
incl
0.81
POSITIVE LOGITS
Small
0.87
small
0.84
ingat
0.81
Large
0.80
Simple
0.80
不仅
0.80
firstly
0.79
Small
0.79
précis
0.78
不僅
0.77
Activations Density 0.175%