INDEX

Explanations

parts of description

The neuron strongly activates on numeric specifications (measurements, quantities, decimals) and the punctuation around them in product or property descriptions.

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

 this

-1.77

 such

-1.62

 that

-1.62

ebly

-1.52

 helping

-1.50

 beziehungs

-1.50

apparence

-1.44

-1.41

もう少し

-1.38

 Ausrüstung

-1.38

POSITIVE LOGITS

 They

1.91

 nozze

1.64

 latine

1.60

 your

1.59

 вашу

1.53

量は

1.47

 มี

1.46

 воспользоваться

1.44

 cerim

1.42

啭

1.41

Activations Density 0.052%