INDEX

Explanations

order

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

 weak

-1.30

weak

-1.14

 Weak

-1.13

 weakness

-1.09

 weakest

-1.06

 weaknesses

-1.00

Weak

-0.96

spots

-0.96

 lemah

-0.95

 weaker

-0.93

POSITIVE LOGITS

 orders

0.69

 ordres

0.59

 ordre

0.50

an

0.47

олові

0.47

orders

0.45

 order

0.45

 orde

0.44

 órdenes

0.41

 при

0.41

Activations Density 0.001%