INDEX

Explanations

code structure and punctuation

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

 would

-1.73

to

-1.68

 still

-1.48

 take

-1.42

get

-1.32

 always

-1.25

 indeed

-1.20

now

-1.20

 ensure

-1.20

 keep

-1.19

POSITIVE LOGITS

 gelegt

1.42

桝

1.38

婊

1.33

 savaş

1.31

 hecha

1.27

咴

1.27

 modernos

1.27

唖

1.26

 reconocidos

1.24

unden

1.23

Activations Density 0.061%