INDEX

Explanations

scientific or medical terms

numeric values in URLs or webpage identifiers.

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

or

-1.70

in

-1.55

 with

-1.45

any

-1.38

on

-1.36

 seemingly

-1.36

 even

-1.34

so

-1.34

 this

-1.33

at

-1.32

POSITIVE LOGITS

 luscious

1.30

 opulent

1.23

 Dinas

1.20

 extravag

1.17

 ayur

1.16

 différents

1.15

 canaux

1.15

particularly

1.14

DONE

1.13

 VERY

1.12

Activations Density 0.000%