INDEX

Explanations

licensed under the Apache license

The neuron primarily detects numeric tokens (digit sequences) and related punctuation (colons, commas, asterisks) such as times, codes, and other number-heavy strings.

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

拟

-0.81

Prescription

-0.72

٤

-0.71

 Lamont

-0.67

 Ван

-0.63

擬

-0.63

alkin

-0.63

 Ramon

-0.62

 במה

-0.62

 scan

-0.61

POSITIVE LOGITS

diksi

0.83

genicity

0.75

定了

0.73

ActivityCompat

0.71

Cancelar

0.68

 Tira

0.68

 Analyst

0.67

matu

0.66

크

0.66

accia

0.66

Activations Density 0.055%