INDEX

Explanations

positive/negative qualities

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

となっています

-1.48

ことがあります

-1.45

our

-1.45

of

-1.43

 Wednesday

-1.41

 dimaksud

-1.41

 mysteriously

-1.36

⃕

-1.36

its

-1.34

 dinosaurio

-1.33

POSITIVE LOGITS

</b>

1.65

</h2>

1.44

"]:

1.43

rounder

1.39

}=\

1.34

eacute

1.33

SetColor

1.25

TokenNameDOT

1.22

',

1.21

Closure

1.20

Activations Density 16.685%