INDEX
Explanations
email addresses in the text
occurrences of the word "Email."
New Auto-Interp
Negative Logits
icle
-0.83
icles
-0.80
osc
-0.68
ICLE
-0.64
steen
-0.64
oscope
-0.63
arist
-0.62
Barth
-0.62
imposed
-0.62
EStreamFrame
-0.61
POSITIVE LOGITS
1.12
0.97
Emails
0.93
0.90
Address
0.89
inbox
0.87
0.79
Subscribe
0.77
legraph
0.76
legram
0.76
Activations Density 0.018%