INDEX

Explanations

if conditions and blocks

The neuron flags conditional keywords and constructs (e.g. “if,” “unless,” “not,” etc.) in code.

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

ଃ

-1.95

ではでは

-1.82

AdvEx

-1.80

 prioridad

-1.73

漚

-1.71

pping

-1.70

絝

-1.70

possibly

-1.69

でご

-1.66

 alábbi

-1.65

POSITIVE LOGITS

فاض

1.66

so

1.50

 thở

1.41

uckleberry

1.36

six

1.36

1.35

近年

1.34

1.33

 epitom

1.33

 stellar

1.31

Activations Density 0.003%