INDEX

Explanations

former/latter states

long sequences

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

ја

0.56

हद

0.45

κά

0.44

非法

0.42

損失

0.42

Lycodon

0.42

ৃতি

0.41

UserID

0.40

ﾔ

0.40

getNumberOf

0.40

POSITIVE LOGITS

 Earlier

0.58

 Formerly

0.57

 ancien

0.54

 تاریخی

0.53

 Modern

0.52

 переменных

0.49

 بیان

0.48

 Perception

0.47

 früheren

0.47

 سابق

0.47

Activations Density 0.000%