INDEX
Explanations
patents and manufacturing
The neuron activates on technical nouns naming the patented object—words like “article,” “sheet,” “membrane,” “preform,” and similar terms that label the invention’s products.
New Auto-Interp
Negative Logits
Still
-0.06
Su
-0.06
слова
-0.06
Receive
-0.06
Cho
-0.06
_accessor
-0.06
Details
-0.06
Native
-0.06
:inline
-0.06
FM
-0.06
POSITIVE LOGITS
trval
0.07
बदल
0.07
bulletin
0.06
�
0.06
konum
0.06
elek
0.06
ология
0.06
SST
0.06
lekker
0.06
ціл
0.06
Activations Density 0.049%