INDEX

Explanations

loop control or code syntax

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

for

-1.52

 régime

-1.34

by

-1.28

 relato

-1.27

 controles

-1.27

all

-1.20

稳定的

-1.19

 szen

-1.17

to

-1.14

-1.13

POSITIVE LOGITS

 allzu

1.33

やう

1.33

たちに

1.23

たびに

1.22

 semejante

1.20

 mètre

1.19

になれば

1.16

 jenen

1.16

It

1.15

áček

1.15

Activations Density 0.001%