INDEX

Explanations

section article citations

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

に行って

-1.06

 estrema

-1.05

えっ

-1.01

œil

-0.96

 bestellt

-0.94

だいぶ

-0.93

 debilidad

-0.93

 мальчик

-0.93

 molta

-0.92

ようです

-0.92

POSITIVE LOGITS

of

3.75

ของ

1.38

 của

1.16

<0xA3>

1.15

вку

1.13

ofthe

1.09

1.07

 ihan

1.06

などの

1.01

а

0.94

Activations Density 0.015%