INDEX

Explanations

of success, risk, or action

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

ed

1.67

Â

1.46

er

1.40

 cardinal

1.30

naam

1.29

<0x0C>

1.29

hearted

1.27

est

1.26

anah

1.26

ামুটি

1.25

POSITIVE LOGITS

自行

2.76

 отклю

2.54

 मोठ्या

2.36

 умень

2.33

 επέ

2.31

 đạt

2.29

芢

2.24

翱

2.23

 również

2.23

占用

2.22

Activations Density 0.100%