INDEX

Explanations

Contact and MS terms found

This neuron responds to standalone numeric tokens and long digit‐ or hex‐like sequences 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

or

-1.35

on

-1.23

-1.17

 Ländern

-1.16

 will

-1.14

 while

-1.12

her

-1.05

-1.02

POSITIVE LOGITS

Öl

1.54

鷽

1.33

футболка

1.28

 nettoyer

1.26

Gön

1.22

Подробнее

1.18

Doktor

1.18

omials

1.18

 coute

1.17

 louer

1.16

Activations Density 0.001%