INDEX

Explanations

brand names and specific product types

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

kle

-0.09

 Pact

-0.09

 Pirate

-0.09

xy

-0.08

 doub

-0.08

 hoop

-0.08

gfx

-0.08

gles

-0.08

jun

-0.08

gro

-0.08

POSITIVE LOGITS

 horrors

0.11

 nests

0.10

 dicks

0.10

naz

0.09

 Dorothy

0.09

mighty

0.09

 careg

0.09

 nich

0.09

 Nich

0.09

reon

0.09

Activations Density 0.003%