INDEX

Explanations

state, locale, cloud

This neuron activates on longer, less common content words (multi-syllable nouns or technical terms).

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

 those

-1.74

 these

-1.47

 affez

-1.44

 your

-1.42

 таких

-1.35

 nemico

-1.34

 ceux

-1.31

 singolo

-1.29

我们

-1.28

لاعات

-1.27

POSITIVE LOGITS

 Kleidung

1.70

in

1.56

 from

1.45



1.41

of

1.38

on

1.37

ské

1.36

 fremden

1.36

 började

1.30

décoration

1.27

Activations Density 0.098%