INDEX

Explanations

dates, and years

The neuron fires on tokens that are numeric sequences—especially those formatted like dates, times, or other digit‐heavy identifiers.

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

in

-1.06

 more

-0.96

ご了承ください

-0.95

 angka

-0.92

 other

-0.90

ご了承

-0.88

for

-0.86

 large

-0.83

星座

-0.82

 biru

-0.81

POSITIVE LOGITS

 ২০১

1.69

 २०

1.39

YYYY

1.34

 YYYY

1.30

yyyy

1.15

 undated

1.05

 vuonna

1.05

 ২০

0.98

 kovo

0.95

 วัน

0.93

Activations Density 0.080%