INDEX

Explanations

before the

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

Negative Logits

0.55

０

0.50

选项

0.50

mán

0.50

、

0.47

teach

0.47

 Authentication

0.46

标识

0.46

 Currie

0.46

 Cupertino

0.46

POSITIVE LOGITS

 herz

0.46

❝

0.46

きちんと

0.46

 protracted

0.45

 consequ

0.44

 physicians

0.43

آم

0.42

 современные

0.42

 dealt

0.42

 herewith

0.42

Activations Density 0.001%