INDEX

Explanations

Biscayne

The neuron fires on tokens that signal “more to come” or continuation/truncation markers (e.g. ellipsis tokens and “Read More »”).

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

you

-1.85

 llena

-1.85

But

-1.75

listdir

-1.72

ărul

-1.70

しかし

-1.64

゛

-1.63

ﺖ

-1.63

لقاء

-1.63

 perpetually

-1.60

POSITIVE LOGITS

1.73

on

1.73

these

1.66

SELF

1.64

itisation

1.63

of

1.62

1.59

感じ

1.57

1.55

媄

1.54

Activations Density 0.000%