INDEX

Explanations

contrasting statements

conversational prompts asking for input

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

 museo

0.52

 लहरा

0.51

 musée

0.50

 সেইরূপ

0.50

 consejo

0.48

 invitación

0.48

accompagn

0.48

 UNESCO

0.47

 Président

0.47

妷

0.47

POSITIVE LOGITS

PRODUCT

0.53

।

0.50

0.47

 breakpoints

0.47

э

0.47

я

0.47

Trusted

0.46

ць

0.46

Height

0.46

Activations Density 0.001%