INDEX

Explanations

properties and identifiers

The neuron fires on occurrences of C#-style property accessors (especially the “get” keyword in property definitions).

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

for

-1.88

if

-1.63

of

-1.59

if

-1.48

our

-1.41

 then

-1.35

-1.30

all

-1.30

-1.27

POSITIVE LOGITS

alers

1.66

⟐

1.65

 reacciones

1.64

考え方

1.52

 jiné

1.50

 mejoras

1.49

 lidé

1.48

Dış

1.47

Ayrıca

1.45

 innych

1.41

Activations Density 0.074%