INDEX

Explanations

even though, even if

introducing specifics after emphasis

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

 tuttavia

0.33

 অথচ

0.32

 Entretanto

0.31

 однако

0.29

 ancak

0.29

 odnosno

0.29

 अतः

0.26

bzw

0.25

 Однако

0.25

°,

0.24

POSITIVE LOGITS

 mores

0.51

 subconsciously

0.40

 acknowledging

0.40

 risking

0.40

 allowing

0.39

 knowing

0.39

 forgetting

0.39

 factoring

0.39

 ignoring

0.38

 hinting

0.38

Activations Density 0.856%