INDEX

Explanations

created by

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

 Moderator

-0.11

-0.10

iene

-0.10

vp

-0.10

 Palin

-0.09

Moder

-0.09

 alma

-0.09

ania

-0.09

 McMahon

-0.09

 shoppers

-0.09

POSITIVE LOGITS

 amateurs

0.18

 skilled

0.15

 inexperienced

0.12

 trained

0.12

 hands

0.12

 amateur

0.11

 professionals

0.11

 qualified

0.11

 someone

0.11

 dile

0.10

Activations Density 0.375%