INDEX

Explanations

end of

The neuron activates on references to endpoints or boundary locations (e.g. “end,” “other end,” “north end,” “south end”).

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

-1.92

Makes

-1.76

-1.55

during

-1.55

arabic

-1.54

vantaged

-1.53

みて

-1.50

”،

-1.42

-1.40

نات

-1.38

POSITIVE LOGITS

end

2.39

ون

2.03

ни

1.85

was

1.80

 Polícia

1.79

1.77

.'

1.68

\&

1.66

鸶

1.64

hysema

1.62

Activations Density 0.048%