INDEX

Explanations

de

The neuron is detecting non-English words (especially Spanish/Portuguese terms with accented characters).

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

-1.84

 parro

-1.78

are

-1.63

-1.62

 recebeu

-1.59

It

-1.54

 メニュー

-1.52

 docteur

-1.50

 Saltar

-1.49

 gräns

-1.47

POSITIVE LOGITS

 your

2.14

your

1.88

 nomor

1.80

 that

1.77

 warszawa

1.73

that

1.69

the

1.68

 parachoque

1.64

暂无

1.63

 year

1.60

Activations Density 0.016%