INDEX

Explanations

country names versus

calculating damage, married women, flushing

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

Ро

0.49

И

0.49

В

0.47

Ко

0.46

Во

0.45

Бо

0.44

આ

0.44

У

0.43

ү

0.43

POSITIVE LOGITS

 sulfide

0.57

 lactate

0.53

 phosphatidyl

0.51

 crows

0.50

 companionship

0.48

 xanth

0.47

ンスター

0.46

qnt

0.46

 भीती

0.46

 cellophane

0.46

Activations Density 0.001%