INDEX

Explanations

programming declarations and specific entities

The neuron activates whenever it sees a declaration keyword (the word “declare”) in code.

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

 anderson

-0.71

IMPORTED

-0.71

erial

-0.70

lichung

-0.68

bullying

-0.66

gist

-0.66

 Nieuws

-0.66

 courteous

-0.65

 Interchange

-0.65

Vita

-0.65

POSITIVE LOGITS

men

0.74

 stimulation

0.74

pb

0.72

mel

0.71

 Decatur

0.70

Observ

0.70

filter

0.70

вых

0.67

stone

0.67

fil

0.65

Activations Density 0.031%