INDEX
Explanations
personal reflection and community
The neuron is sensitive to temporal and duration words—tokens that refer to time (e.g. “time,” “year,” “week,” “month,” “together”), marking when or how long events occur.
New Auto-Interp
Negative Logits
terrorists
-0.08
Mat
-0.08
seals
-0.07
cross
-0.07
noise
-0.07
eski
-0.07
IPP
-0.07
feat
-0.07
Struct
-0.07
Sonia
-0.06
POSITIVE LOGITS
acle
0.06
VERTISEMENT
0.06
'})
0.06
licher
0.05
завд
0.05
.build
0.05
čtvrt
0.05
.iloc
0.05
것을
0.05
blob
0.05
Activations Density 0.138%