INDEX

Explanations

technical terms and French words

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 занятия

1.73

los

1.72

ne

1.70

こそ

1.67

ѐ

1.67

জায়

1.64

өм

1.64

 delito

1.62

 giocatore

1.62

ሉት

1.61

POSITIVE LOGITS

ოს

1.69

 implant

1.46

ycznych

1.46

 situ

1.44

 sketched

1.43

 expedient

1.42

 odred

1.40

 WASH

1.39

tand

1.39

 tang

1.39

Activations Density 0.000%

No Known Activations