INDEX

Explanations

publishers and books

The neuron strongly activates on the names of publishing houses and imprints (e.g. Wiley, Springer, Pantheon).

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

 that

-1.77

and

-1.54

on

-1.33

 which

-1.27

if

-1.23

our

-1.16

an

-1.15

んですよ

-1.14

 about

-1.09

の方に

-1.08

POSITIVE LOGITS

<bos>

1.45

がありますが

1.39

 nové

1.38

 flera

1.30

 няколко

1.26

 récentes

1.24

 ersche

1.24

ただし

1.23

 eredeti

1.21

 когато

1.20

Activations Density 0.003%