INDEX

Explanations

publicly available and approved

consequence of actions

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 Iran

0.47

ro

0.46

 अशा

0.42

 dado

0.41

 Senator

0.40

 visitation

0.40

ఘ

0.40

 Appro

0.40

lement

0.39

western

0.39

POSITIVE LOGITS

 APPEAL

0.55



0.50

＋

0.49

MIS

0.49

FIT

0.49

ACES

0.48

EFF

0.47

 weep

0.47

♯

0.47

 conséquent

0.46

Activations Density 0.001%