INDEX
Explanations
technical writing
The neuron primarily activates on common function words (e.g. articles, prepositions, and other grammatical particles).
New Auto-Interp
Negative Logits
fists
-0.07
_subscription
-0.07
vd
-0.07
Videos
-0.06
.documents
-0.06
ceased
-0.06
bsites
-0.06
tracts
-0.06
SSID
-0.06
Subset
-0.06
POSITIVE LOGITS
deltaX
0.06
Bergen
0.06
Amp
0.06
(cid
0.06
ένας
0.06
}}
0.06
�
0.06
})(
0.06
�
0.06
bagi
0.06
Activations Density 0.059%