INDEX
Explanations
This neuron detects the standalone indefinite article “a” in the text.
New Auto-Interp
Negative Logits
Kernel
-0.07
상의
-0.06
گران
-0.06
Thing
-0.06
подоб
-0.06
kernel
-0.06
Index
-0.06
differs
-0.06
کس
-0.06
_ind
-0.06
POSITIVE LOGITS
Sexual
0.07
imageURL
0.07
Enhanced
0.06
rompt
0.06
encompass
0.06
Date
0.06
Rotary
0.06
/as
0.06
ComVisible
0.06
actus
0.06
Activations Density 0.008%