INDEX

Explanations

tools or materials

The neuron activates on words naming tools, equipment, or implements (e.g. socket, ratchet, drill, sandpaper).

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

ݮ

-1.54

 maniere

-1.49

 brune

-1.43

Tentang

-1.42

 avocat

-1.41

 carreaux

-1.41

 conster

-1.41

 plafon

-1.39

㈬

-1.38

 appels

-1.38

POSITIVE LOGITS

or

3.13

if

2.66

of

2.13

to

1.89

 then

1.74

in

1.70

 like

1.63

 other

1.62

或

1.59

 either

1.57

Activations Density 0.066%