INDEX

Explanations

substitution and subsequent actions

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

datable

-1.62

 silicona

-1.59

führen

-1.53

BACA

-1.52

 arbe

-1.49

 direkte

-1.48

 DialogInterface

-1.48

はこの

-1.46

他的

-1.45

dwa

-1.41

POSITIVE LOGITS

for

2.02

our

1.89

 into

1.85

1.67

the

1.58

1.50

 entanto

1.50

to

1.48

 fofo

1.45

 greater

1.42

Activations Density 0.020%