INDEX

Explanations

specific proper nouns

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

 another

-1.30

 what

-1.30

 when

-1.11

另一个

-1.10

if

-1.09

one

-1.04

 before

-1.01

our

-1.00

czego

-1.00

另一

-0.97

POSITIVE LOGITS

the

1.62

The

1.45

They

1.20

 وذلك

1.09

 Vereinig

1.09

The

1.04

six

1.00

five

1.00

lor

0.99

printStackTrace

0.99

Activations Density 0.097%