INDEX

Explanations

App specifications and status

The neuron fires on the capitalized “keys” or field‐name labels (e.g. Name, Version, Release, Languages, Developer, Height, Weight, etc.) in structured metadata lists.

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

gilet

-1.59

and

-1.57

 nosso

-1.50

our

-1.45

 occasione

-1.36

 nuestros

-1.34

larımız

-1.33

 when

-1.27

 Ähn

-1.25

 quando

-1.23

POSITIVE LOGITS

1.48

 ご飯

1.30

its

1.30

 també

1.27

匱

1.26

ệu

1.23

 також

1.23

七月

1.22

 vong

1.20

 zároveň

1.20

Activations Density 0.335%