INDEX

Explanations

decentralized, deactivations, de-escalation, deception

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

حاء

1.93

торое

1.85

เที่

1.83

 Viitattu

1.81

ץ

1.80

dala

1.78

ęs

1.77

リューム

1.72

iquity

1.72

streng

1.72

POSITIVE LOGITS

 facto

1.61

지털

1.50

utsches

1.36

inition

1.34

uterte

1.33

тельность

1.32

isive

1.32

utscher

1.31

ceased

1.29

구나

1.28

Activations Density 0.164%