INDEX

Explanations

in `List` or `(`)

The neuron is picking out numeric literals (numbers) in the code.

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

-2.09

-1.62

-1.57

,'

-1.44

-1.43

-1.38

 Она

-1.37

[],

-1.34

$\

-1.32

 связанные

-1.32

POSITIVE LOGITS

 aceasta

1.74

henge

1.68

 persoane

1.66

 urme

1.63

'';

1.61

 personalización

1.60

“-

1.54

 tracción

1.53

’).

1.53

"-"

1.50

Activations Density 0.038%