INDEX

Explanations

cannot fulfill requests

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

iž

1.20

マスク

1.12

خ

1.11

avalt

1.10

規制

1.09

控制器

1.09

uncu

1.09

交叉

1.08

сток

1.08

чиков

1.08

POSITIVE LOGITS

 railroads

1.23

 kana

1.17

cil

1.17

aaa

1.15

 hover

1.14

 time

1.14

LEM

1.12

 illa

1.11

 Wrong

1.11

 Technically

1.10

Activations Density 0.013%