INDEX

Explanations

one and the other

The neuron fires on numeric tokens (integers) occurring in the text.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 even

-1.23

it

-1.22

 TAMBÉM

-1.21

Кроме

-1.13

 również

-1.12

 også

-1.08

certain

-1.07

 även

-1.05

甚至

-1.04

甚至是

-1.04

POSITIVE LOGITS

and

1.03

 then

1.02

 andando

1.00

 conversely

0.95

 دیگری

0.93

หนึ่ง

0.91

burgs

0.91

while

0.86

 while

0.85

 resor

0.84

Activations Density 0.088%