INDEX

Explanations

copyright and dimensions

The neuron fires on numeric tokens—especially sequences of digits (often with decimals) used in technical dimensions, measurements, or version/license numbers.

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

 Forse

-0.84

 キャラクター

-0.82

 artig

-0.80

📉

-0.78

 nominate

-0.77

isFetching

-0.77

TÜ

-0.76

 serangga

-0.74

vœ

-0.74

 highlighting

-0.73

POSITIVE LOGITS

ORD

0.84

 Dimension

0.83

iced

0.82

shear

0.81

 ซี

0.80

Dimension

0.80

/>

0.79

huber

0.76

род

0.76

priate

0.76

Activations Density 0.008%