INDEX

Explanations

contextual items and descriptions

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

all

-1.84

で一番

-1.49

gesetzt

-1.45

 also

-1.40

 others

-1.34

 that

-1.30

 each

-1.28

not

-1.26

 everybody

-1.26

 afterwards

-1.25

POSITIVE LOGITS

 هذا

1.37

 может

1.35

хіво

1.28

ETHODS

1.21

it

1.20

 padx

1.18

 puede

1.17

::$_

1.17

 debe

1.16

 توانید

1.16

Activations Density 0.413%