INDEX

Explanations

starts with 'Subject:'

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

thirty

1.26

PLATES

1.25

zsche

1.24

月末

1.21

十九章

1.20

 Thirty

1.18

twenty

1.18

ruary

1.17

開幕

1.16

ILLIPS

1.16

POSITIVE LOGITS

 goin

1.61

たい

1.55

 puedes

1.53

 poderia

1.50

CLUSTERED

1.49

 Feld

1.46

 проє

1.46

Needs

1.44

誤差

1.43

 بیمار

1.42

Activations Density 0.047%