INDEX

Explanations

explaining these points

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

to

1.32

by

1.31

umb

1.27

álně

1.17

al

1.17

for

1.16

in

1.16

mult

1.14

laş

1.13

bym

1.09

POSITIVE LOGITS

Enlaces

1.57

 aquello

1.47

Bra

1.43

<unused2059>

1.42

Spare

1.40

 cuchar

1.40

oLocations

1.38

라이언

1.38

りの

1.37

EnglishMarks

1.36

Activations Density 1.012%