INDEX

Explanations

a, an

New Auto-Interp

Configuration

Prompts (Dashboard)

16,384 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

parsedMessage

-0.75

openConnection

-0.73

RenderAtEndOf

-0.71

principalTable

-0.65

abestanden

-0.62

Бахар

-0.60

kháu

-0.60

 simplifié

-0.58

 barbarians

-0.57

 Hickman

-0.57

POSITIVE LOGITS

Rüyada

0.58

 balle

0.49

аза

0.48

FormTagHelper

0.48

ⓧ

0.46

AndEndTag

0.45

 nooit

0.43

ględ

0.43

 concerned

0.43

 expected

0.42

Activations Density 0.009%