INDEX

Explanations

true or false values

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

of

-2.11

 when

-1.45

why

-1.33

 after

-1.30

 because

-1.23

 Сла

-1.20

how

-1.20

 seguirá

-1.20

 whose

-1.19

 Auffassung

-1.19

POSITIVE LOGITS

țial

1.55

 "");

1.36

keiten

1.27

ι

1.23

without

1.23

culoskeletal

1.22

TargetException

1.22

iniums

1.21

észetes

1.20

paedia

1.20

Activations Density 0.015%