INDEX

Explanations

controlled or managed 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

谁能

-1.78

 अच्छा

-1.58

 errori

-1.51

𐌽

-1.51

étaires

-1.51

 cioc

-1.49

gruss

-1.47

 indietro

-1.45

λή

-1.45

ᮣ

-1.42

POSITIVE LOGITS

to

1.59

ed

1.48

1.42

."

1.41

We

1.39

1.34

and

1.34

 наших

1.34

随意

1.29

—

1.26

Activations Density 0.138%