variable names followed by comparison

The neuron flags the program variable (identifier) being checked on the left‐hand side of an if/elif conditional.

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

 bolsos

-1.63

 Canadá

-1.62

ırlar

-1.60

 univerz

-1.59

 camisas

-1.55

 validés

-1.55

 luzes

-1.54

 disponibilités

-1.52

 espere

-1.51

anız

-1.49

POSITIVE LOGITS

1.59

1.53

1.52

ra

1.52

1.45

ne

1.41

Activations Density 0.002%