INDEX

Explanations

discussing or asking about

The neuron strongly activates on occurrences of the word “about.”

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

);

-1.74

-1.63

 trusty

-1.56

*}{

-1.55

in

-1.52

),

-1.51

 Those

-1.49

For

-1.48

[]:

-1.48

 awful

-1.47

POSITIVE LOGITS

how

1.78

茈

1.55

 baumwolle

1.51

هو

1.48

 pérd

1.45

 bescher

1.44

 INNOV

1.38

RÉ

1.38

 みんな

1.37

 glyphicon

1.34

Activations Density 0.053%