INDEX
Explanations
conversations, specifically containing phrases such as "Hey man, we wanna make this into a movie"
informal conversational phrases and dialogue
New Auto-Interp
Negative Logits
etheless
-0.82
prisingly
-0.80
uitive
-0.77
surprisingly
-0.76
minist
-0.75
respondents
-0.67
mittedly
-0.66
cture
-0.60
arten
-0.59
:=
-0.59
POSITIVE LOGITS
.")
1.71
").
1.62
"]
1.46
")
1.45
"),
1.44
!".
1.40
'"
1.37
)"
1.37
)",
1.35
)."
1.35
Activations Density 0.786%