INDEX

Explanations

selecting an answer

evaluating states and probabilities

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

ിച്ച്

0.46

একটি

0.44

 എന്നാണ്

0.43

 ขณะ

0.43

 وهذا

0.42

 ವಾ

0.42

 അപകട

0.41

のも

0.40

 Ваш

0.39

 USING

0.39

POSITIVE LOGITS

 magician

0.50

 fels

0.45

egin

0.44

uris

0.43

飙

0.43

 mandala

0.42

 perfection

0.41

 milling

0.41

 mandarin

0.40

ném

0.40

Activations Density 0.000%