INDEX

Explanations

code property definitions

The neuron fires on numeric literal tokens (e.g. numbers appearing in code assignments or property initializers).

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

我只

-1.77

我只是

-1.74

).__

-1.59

]:

-1.55

«

-1.53

—

-1.52

我就是

-1.50

サート

-1.49

要不是

-1.48

__).

-1.46

POSITIVE LOGITS

 These

1.86

 această

1.66

 aceste

1.60

这些

1.59

 comenzar

1.57

Etimología

1.57

 Accueil

1.55

這個

1.54

 suoi

1.53

↵↵

1.52

Activations Density 0.004%