INDEX
Explanations
most, many, some followed by a noun
New Auto-Interp
Negative Logits
他人
0.46
Others
0.44
decenas
0.44
portions
0.42
Others
0.41
относя
0.41
others
0.41
ევრი
0.40
aspectos
0.40
aspects
0.40
POSITIVE LOGITS
newer
0.78
larger
0.68
bigger
0.63
older
0.59
större
0.59
newer
0.56
nicer
0.55
nowadays
0.55
thicker
0.54
larger
0.54
Activations Density 0.035%