INDEX

Explanations

determiners followed by nouns

This neuron fires strongly on capitalized words beginning new sentences or sections (i.e. sentence-initial tokens).

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

攫

-1.62

 början

-1.50

 what

-1.42

 ricas

-1.34

ların

-1.34

キニ

-1.34

冽

-1.34

。】

-1.34

咵

-1.33

its

-1.31

POSITIVE LOGITS

you

1.73

 هذه

1.55

are

1.46

也要

1.36

self

1.34

 queste

1.29

of

1.27

 these

1.23

 aceasta

1.22

on

1.21

Activations Density 0.079%