INDEX

Explanations

chauffeDaunrise, Nonetheless, Noriega

The neuron fires on numeric tokens—especially decimal or measured values—i.e. it detects numbers in the text.

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

but

-2.22

↵↵

-1.74

但在

-1.63

 scuro

-1.53

だったが

-1.52

ordinator

-1.48

 constituye

-1.47

‌.

-1.46

感じる

-1.46

睹

-1.43

POSITIVE LOGITS

𓆟

1.98

 OGSÅ

1.52

廍

1.48

他にも

1.43

ウザ

1.41

<b>

1.40

ならではの

1.39

 chauffe

1.38

 them

1.37

 tombé

1.36

Activations Density 0.002%