INDEX

Explanations

code imports and types

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

in

-1.55

 with

-1.50

and

-1.33

anxi

-1.30

indiri

-1.30

to

-1.26

by

-1.25

at

-1.24

all

-1.23

corollary

-1.20

POSITIVE LOGITS

lizing

1.34

ulation

1.34

fies

1.33

ISED

1.33

isers

1.33

 corrom

1.32

tical

1.30

ofthe

1.30

 turco

1.29

itizing

1.29

Activations Density 0.169%