INDEX

Explanations

through methods and processes

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

 worse

1.26

 wasn

1.11

 estava

1.10

for

1.10

 Worse

1.10

 president

1.08

ute

1.05

was

1.04

 für

1.03

 busier

1.03

POSITIVE LOGITS

 ذریع

1.46

自身的

1.39

อุปกรณ์

1.38

豊富な

1.31

自身の

1.30

ならでは

1.26

ських

1.25

独自の

1.20

 svojih

1.19

ศิลป

1.19

Activations Density 1.286%