INDEX

Explanations

slice or cross section

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

ness

-0.07

ibal

-0.07

Ĺ

-0.07

-dom

-0.07

loo

-0.07

owers

-0.07

anium

-0.07

éĹªéĹª

-0.07

iao

-0.06

pris

-0.06

POSITIVE LOGITS

slice

0.11

 slices

0.11

 slice

0.10

(slice

0.09

åīĸéĿ¢

0.09

 Slice

0.09

 perpendicular

0.09

 slicing

0.08

napshot

0.08

snap

0.08

Activations Density 0.028%