INDEX

Explanations

not modifying the original

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

 hadn

1.07

 recognizes

1.05

 recognized

1.03

 recognize

1.03

 obvious

0.98

 wasn

0.98

foc

0.98

 για

0.98

 ਲਈ

0.97

alo

0.95

POSITIVE LOGITS

oscopic

1.41

,*/}*.

1.29

фа

1.26

quiz

1.23

LINK

1.23

,*/}

1.20

CUR

1.19

ंत्रिक

1.14

 }^{\

1.14

subequations

1.13

Activations Density 0.008%