INDEX

Explanations

step by step explanation

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 lakini

1.91

роў

1.85

しますが

1.79

раў

1.78

そうで

1.75

นะครับ

1.72

ayanti

1.72

帮你

1.70

ǎo

1.64

あなたの

1.63

POSITIVE LOGITS

led

1.46

 unique

1.27

 smoked

1.26

kul

1.25

Unique

1.24

 guts

1.22

 implied

1.21

Led

1.20

 dreamed

1.20

 dreamt

1.18

Activations Density 0.001%