INDEX

Explanations

technical/informational content

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 myſelf

-1.84

 itſelf

-1.71

 Theſe

-1.70

 pleaſure

-1.65

Efq

-1.63

 Monfieur

-1.63

 ſeveral

-1.60

 ſtate

-1.55

 Majefty

-1.55

 reaſon

-1.55

POSITIVE LOGITS

1.27

of

1.26

1.24

1.22

in

1.20

and

1.20

for

1.20

1.19

1.07

↵↵

1.04

Activations Density 0.181%