INDEX

Explanations

This neuron detects numeric tokens or sequences of digits (e.g., multi‐digit numbers) in the text.

time and speed metrics

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

 this

-1.54

 these

-1.36

енча

-1.33

 ваз

-1.27

 important

-1.26

ができます

-1.22

 that

-1.22

 will

-1.20

 could

-1.19

 виправи

-1.18

POSITIVE LOGITS

 nových

1.25

ſo

1.23

 egreg

1.23

 thậm

1.23

 Unterrichts

1.16

 Dinas

1.15

 Inſ

1.14

̏

1.13

 angebracht

1.12

ныхъ

1.10

Activations Density 0.010%