INDEX

Explanations

persuading, convincing, counseling

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 such

-1.56

Zubereitung

-1.35

-1.30

llavero

-1.30

Więcej

-1.28

 prominently

-1.28

we

-1.27

Ꮞ

-1.25

𞤢

-1.24

dados

-1.23

POSITIVE LOGITS

You

1.52

why

1.52

Thus

1.51

ipzig

1.45

 dlaczego

1.44

糴

1.44

énieur

1.42

,”

1.41

dders

1.40

sizeCache

1.40

Activations Density 0.019%

No Known Activations