INDEX

Explanations

creative expression and problem solving

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 deliberate

-0.11

onis

-0.10

endum

-0.10

lev

-0.09

oby

-0.09

avier

-0.09

-tone

-0.09

ÑĢÐ¾Ñģ

-0.09

Rand

-0.09

st

-0.09

POSITIVE LOGITS

 Commons

0.16

 juices

0.14

 commons

0.12

/original

0.12

commons

0.12

/art

0.12

 flair

0.10

 Creat

0.10

inine

0.10

-destruct

0.10

Activations Density 0.033%

No Known Activations