INDEX

Explanations

steps lists

applied cables smaller grain

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

’

0.56

 starts

0.43

 surf

0.42

 Motel

0.42

 recycling

0.40

<unused60>

0.39

 diners

0.39

 diner

0.39

oring

0.38

 surfing

0.38

POSITIVE LOGITS

 genü

0.48

 karşınız

0.47

 zugleich

0.47

 પહો

0.46

 depl

0.46

 könne

0.45

 entgegen

0.44

 asemenea

0.44

 compless

0.44

 كرد

0.44

Activations Density 0.000%