INDEX

Explanations

explaining causes or reasons

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

or

-1.64

 using

-1.40

 with

-1.40

 utilizando

-1.23

by

-1.20

 plus

-1.20

 including

-1.19

to

-1.16

 based

-1.14

 without

-1.10

POSITIVE LOGITS

 because

2.25

因为它

1.78

 потому

1.72

因为

1.64

 puisqu

1.60

ecause

1.57

因為

1.56

 porque

1.53

 karena

1.51

 since

1.38

Activations Density 0.227%