INDEX

Explanations

names followed by surnames or initials

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

aN

-0.11

 Islam

-0.09

ini

-0.09

 assembly

-0.08

(@

-0.08

 Islamist

-0.08

 Cancel

-0.08

 Hernandez

-0.08

@js

-0.08

 manus

-0.08

POSITIVE LOGITS

ueil

0.09

odiac

0.09

 Synd

0.09

Ã¼ss

0.08

 Sydney

0.08

 Sidney

0.08

phil

0.08

 detach

0.08

eltas

0.08

 dies

0.08

Activations Density 0.338%