INDEX

Explanations

midpoint list division

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

тай

2.49

to

2.40

 ολο

2.30

 تبدیل

2.26

monster

2.25

 레드

2.21

much

2.18

bý

2.14

serve

2.13

יו

2.13

POSITIVE LOGITS

eea

2.94

erent

2.94

ה

2.88

🥯

2.85

\{

2.78

тво

2.70

ructuring

2.70

ece

2.68

ei

2.68

crumb

2.66

Activations Density 0.018%