INDEX
Explanations
structured data in a tabular format
New Auto-Interp
Negative Logits
]âĢı
-0.15
l
-0.15
1
-0.15
j
-0.15
c
-0.15
رش
-0.14
z
-0.14
t
-0.14
3
-0.13
Twenty
-0.13
POSITIVE LOGITS
0.27
NaN
0.19
NaN
0.17
a
0.17
etc
0.17
b
0.16
s
0.16
p
0.16
nan
0.16
g
0.15
Activations Density 0.018%