INDEX

Explanations

for loop indices

The neuron detects numeric literal tokens (numbers) in code snippets.

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

 גם

-1.27

 także

-1.26

 også

-1.22

 επίσης

-1.19

utdown

-1.17

 také

-1.16

 نیز

-1.15

 perca

-1.14

vação

-1.14

 также

-1.14

POSITIVE LOGITS

 those

1.47

 existentes

1.39

 Lengkap

1.33

 privadas

1.33

 existing

1.30

italien

1.27

 Anerkennung

1.25

 Nachfrage

1.25

 exteriores

1.25

it

1.25

Activations Density 0.003%