INDEX

Explanations

called, named, or called it

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

 Aero

-0.09

Tol

-0.09

arto

-0.08

 infl

-0.08

 Liberties

-0.08

uben

-0.07

igel

-0.07

GT

-0.07

POSITIVE LOGITS

 gener

0.15

 joking

0.12

 collo

0.12

 locally

0.12

lam

0.11

 collectively

0.11

 errone

0.11

eup

0.11

 vern

0.11

 interchange

0.11

Activations Density 0.072%