INDEX

Explanations

and followed by various words

and followed by negative or contrasting ideas

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

 ovales

0.26

 fillets

0.24

 five

0.23

 sunglasses

0.22

 antibiotics

0.22

 alphanumeric

0.22

 browsers

0.22

 handicrafts

0.22

 manuals

0.22

 buttons

0.22

POSITIVE LOGITS

 понима

0.24

 справедливо

0.24

 unwillingness

0.22

 bukanlah

0.22

 downright

0.22

 sufrimiento

0.22

 disrespect

0.21

 indign

0.21

 όχι

0.21

 способности

0.21

Activations Density 1.065%