INDEX

Explanations

numbers following punctuation

This neuron fires on numeric tokens—especially multi-digit numbers like years, dates, and numeric references.

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

頃に

-1.49

 Italijani

-1.46

いますが

-1.45

 asesinado

-1.41

不清楚

-1.41

combined

-1.40

 تواند

-1.38

 médec

-1.38

 adecuados

-1.37

 entraîner

-1.37

POSITIVE LOGITS

 every

1.78

all

1.73

 each

1.65

 those

1.52

ทุก

1.38

 their

1.37

 COMERC

1.34

︱

1.30

 these

1.25

any

1.20

Activations Density 0.017%