INDEX

Explanations

news updates

hidden magical assessments isolation

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

σετε

0.47

せた

0.47

 obter

0.46

 croche

0.45

ран

0.44

くない

0.44

 chở

0.43

щую

0.43

щем

0.43

디오

0.42

POSITIVE LOGITS

ali

0.57

be

0.55

0.54

কিন্তু

0.53

frü

0.53

ulos

0.51

all

0.51

me

0.51

YM

0.50

 विराजमान

0.50

Activations Density 0.001%