INDEX

Explanations

Builder pattern

This neuron fires on programming identifiers written in PascalCase (e.g. class, constructor, or provider names).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 Popis

-1.39

 Sünden

-1.33

-1.30

 migliorare

-1.23

فیلم

-1.23

 telefónica

-1.20

ouac

-1.19

 такими

-1.19

 stieg

-1.19

aporation

-1.18

POSITIVE LOGITS

 provar

1.46

 PHOSPH

1.39

of

1.37

髕

1.35

 계획

1.30

 presenti

1.28

ditionally

1.28

的后

1.25

 Linkage

1.25

 continuo

1.23

Activations Density 0.038%