INDEX

Explanations

various components or services

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

any

-1.64

 even

-1.11

 actually

-1.06

 really

-1.05

 anything

-1.02

 their

-1.00

 just

-0.96

at

-0.96

 starts

-0.95

all

-0.94

POSITIVE LOGITS

Various

1.37

various

1.30

Certain

1.22

 некоторы

1.20

several

1.18

 birçok

1.18

sometimes

1.18

 parfois

1.17

 некоторых

1.16

Occasionally

1.16

Activations Density 0.135%