INDEX

Explanations

step by

The neuron consistently highlights standalone numeric tokens (especially single digits like “0,” “1,” “2,” etc.), effectively detecting numbers 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

 what

-1.58

 will

-1.37

 this

-1.36

おわりに

-1.24

its

-1.23

For

-1.20

 that

-1.18

 That

-1.10

 into

-1.03

 Once

-1.03

POSITIVE LOGITS



1.25

Jing

1.18

உ

1.17

Αν

1.15

reonine

1.13

こちらは

1.12

これは

1.12

 pien

1.11

馓

1.10

 Champaign

1.08

Activations Density 0.005%