INDEX

Explanations

using namespace and html tags

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

coled

-1.59

𓆏

-1.58

senic

-1.46

𓍊

-1.44

orale

-1.41

珰

-1.40

耵

-1.38

 nowa

-1.37

 jesteśmy

-1.36

gallo

-1.34

POSITIVE LOGITS

 where

1.80

 when

1.77

↵↵

1.77

↵↵↵

1.55

 with

1.38

1.34

to

1.29

for

1.26

1.23

 видов

1.22

Activations Density 0.044%