INDEX

Explanations

events and activities

The main thing this neuron does is detect words referring to creative or artistic activities (like “producing,” “creating,” “mix,” “events,” “works,” “plays”).

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

in

-2.36

 there

-1.68

it

-1.51

at

-1.44

 even

-1.41

 then

-1.38

 this

-1.32

the

-1.30

you

-1.26

↵↵↵

-1.23

POSITIVE LOGITS

ställ

1.42

föra

1.27

ässe

1.26

ställning

1.26

 appunto

1.23

 Hivatalos

1.22

@{

1.21

ᥕ

1.21

BOTH

1.19

ALLES

1.19

Activations Density 0.171%