INDEX

Explanations

urls and domain names

digits within dates, timestamps, and numerical identifiers.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

</em>

-2.33

 дочь

-2.13

 этом

-2.08

 нашу

-2.08

 сына

-2.02

 prace

-2.00

 этого

-2.00

 美味

-1.99

 لعبة

-1.97

 изпол

-1.96

POSITIVE LOGITS

in

2.95

if

2.53

the

2.52

at

2.41

by

2.08

1.98

 ilumina

1.96

on

1.80

 ofrecen

1.77

it

1.77

Activations Density 0.058%