INDEX

Explanations

offices, forums, boards

The neuron fires on forum interface/header words (e.g. “forums,” “boards,” “You are currently viewing our boards,” etc.).

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

and

-1.91

них

-1.51

νες

-1.41

kých

-1.31

RICULUM

-1.29

 अन्य

-1.26

 contextes

-1.23

撻

-1.23

muting

-1.23

[...,

-1.23

POSITIVE LOGITS

 what

1.48

1.45

着一个

1.41

how

1.38

 Haare

1.34

 основных

1.31

 Miß

1.27

 Probleme

1.24

iniums

1.20

 Initially

1.18

Activations Density 0.188%