INDEX

Explanations

the letter U or overview

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

面

1.44

Buc

1.27

aer

1.24

кта

1.23

 tere

1.22

 conceal

1.22

粉

1.20

lef

1.19

 resol

1.18

 гли

1.18

POSITIVE LOGITS

 dola

1.61

{\$

1.55

stvo

1.48

impero

1.47

 anders

1.46

 scienza

1.44

 torno

1.44

wiad

1.43

 abuso

1.42

 terang

1.40

Activations Density 0.003%