INDEX

Explanations

match or matching color

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

ahu

-0.11

laus

-0.10

 pref

-0.10

pot

-0.10

lav

-0.09

 lens

-0.09

_based

-0.09

 Sloan

-0.09

 Jefferson

-0.09

Sak

-0.08

POSITIVE LOGITS

eam

0.11

estar

0.10

 dubious

0.09

leh

0.09

venes

0.09

 ident

0.09

 Knee

0.09

duk

0.09

 perfectly

0.09

 Pocket

0.09

Activations Density 0.014%