INDEX

Explanations

introducing a statement or consequence

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

inson

-0.08

 graz

-0.07

 wisely

-0.07

buch

-0.07

loo

-0.07

.Ar

-0.07

 luckily

-0.07

 lucky

-0.07

 Prefer

-0.07

annels

-0.07

POSITIVE LOGITS

adf

0.11

ONSE

0.10

 does

0.10

iew

0.10

Ø¶

0.10

do

0.10

too

0.10

ady

0.10

iten

0.10

&(

0.10

Activations Density 0.049%

No Known Activations