INDEX
Explanations
Instructions
The neuron detects instructions telling the reader to include or reference specific keywords (e.g. put/note/reference “…” on the memo or subject line).
New Auto-Interp
Negative Logits
BM
-0.06
補
-0.06
Fallout
-0.06
derby
-0.06
endo
-0.06
补
-0.06
trackers
-0.06
vend
-0.06
bos
-0.06
ensemble
-0.06
POSITIVE LOGITS
prioritize
0.07
지를
0.07
Brazil
0.06
ursos
0.06
[data
0.06
Mean
0.06
*this
0.06
trước
0.06
(Event
0.06
스의
0.06
Activations Density 0.015%