INDEX
Explanations
turn to phrases and prompts
New Auto-Interp
Negative Logits
Olimpi
0.84
accumulation
0.82
ومن
0.82
ammer
0.81
Panoramic
0.80
منهج
0.79
土産
0.78
Découvrez
0.78
Referências
0.77
<unused1639>
0.77
POSITIVE LOGITS
prompts
0.80
prompted
0.72
ishing
0.68
prompt
0.68
cor
0.67
cl
0.66
стные
0.64
oh
0.63
heads
0.63
prompt
0.62
Activations Density 0.000%