INDEX

Explanations

budgets

descriptions followed by continuations

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 Forsyth

0.48

strual

0.47

 Laus

0.46

gif

0.44

MFA

0.43

ո

0.43

gfx

0.42

.');

0.42

serializer

0.41

opausal

0.41

POSITIVE LOGITS

TEL

0.59

BH

0.58

ברו

0.57

0.54

SCI

0.53

TAL

0.53

ैल

0.52

BUR

0.52

TEC

0.50

Bren

0.50

Activations Density 0.000%