INDEX

Explanations

path attribute value

The neuron activates specifically on the identifier “path” in code, especially in route definitions or attribute assignments.

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

by

-2.14

-1.96

 also

-1.87

 other

-1.65

 інші

-1.61

for

-1.56

 still

-1.54

add

-1.52

了一会儿

-1.52

ab

-1.46

POSITIVE LOGITS



1.66

bilden

1.48

 значит

1.48

 mulheres

1.48

MINUTE

1.42

følge

1.41

OnPage

1.41

 maravill

1.40

稣

1.38

NEY

1.36

Activations Density 0.001%