INDEX

Explanations

return prime

The neuron activates on numeric literals and related tokens (constants, numeric values, and their signs).

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 коро

0.56

 Pflicht

0.55

衝

0.55

 getty

0.53

 feathered

0.52

 vidé

0.50

 Getty

0.49

맞

0.49

 leichte

0.49

 герцо

0.48

POSITIVE LOGITS

]=-

1.27

=-

1.27

(-

1.23

=-

1.20

(-

1.19

==-

1.17

}=-

1.13

,-

1.12

)=-

1.12

)(-

1.07

Activations Density 0.010%