INDEX

Explanations

untreated subjects or conditions

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

-2.61



-2.53

 แต่

-2.36

佧

-2.34

変わる

-2.28

 étrangères

-2.25

来た

-2.22

亖

-2.17

всегда

-2.11

 tuvo

-2.09

POSITIVE LOGITS

ка

3.11

↵

2.94

酲

2.48

 downloading

2.44

 handing

2.42

 coordinating

2.33

 utilising

2.30

 evaluating

2.25

曩

2.25

 implementing

2.23

Activations Density 0.004%