INDEX

Explanations

fully interchangeable

contrasting information or qualifications

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

 refueling

0.51

 वर

0.49

WORKING

0.48

ল

0.47

 tariffs

0.47

 utilisée

0.47

挞

0.46

を使って

0.46

廃棄

0.45

 deformation

0.45

POSITIVE LOGITS

0.59

ved

0.57

lll

0.57

bm

0.55

ih

0.55

zed

0.55

cken

0.54

mn

0.54

bbe

0.53

nej

0.53

Activations Density 0.003%