INDEX

Explanations

following specific words

structure of a

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

'[

0.44

 jamming

0.44

."

0.42

 ischa

0.42

 spreadsheet

0.40

mnop

0.40

 comments

0.40

臥

0.39

 versions

0.39

 refers

0.39

POSITIVE LOGITS

 เสียง

0.50

SIDE

0.47

오늘은

0.45

 entusiasmo

0.44

rewards

0.44

┈

0.44

reward

0.44

 基本

0.44

ТА

0.44

 aște

0.43

Activations Density 0.001%