INDEX

Explanations

application for

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Dul

-0.12

 Dahl

-0.11

 enumerated

-0.10

 Wong

-0.10

 MacDonald

-0.10

zet

-0.10

(strip

-0.09

vt

-0.09

 herein

-0.09

wit

-0.09

POSITIVE LOGITS

 relent

0.10

 illness

0.10

 pals

0.09

Kop

0.09

 Fever

0.09

 fever

0.09

 craz

0.09

 stealth

0.08

kur

0.08

 altered

0.08

Activations Density 0.258%