INDEX

Explanations

signal correction and equalization

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

ская

-1.71



-1.50

 Ankunft

-1.49

ⓕ

-1.44

"));

-1.42

黩

-1.42

 decât

-1.41

好像是

-1.41

達が

-1.40



-1.39

POSITIVE LOGITS

1.42

 Steer

1.38

 Piles

1.33

睃

1.28

 quoque

1.23

 tirantes

1.22

utivos

1.22

你和

1.20

و

1.19

轵

1.19

Activations Density 0.015%