INDEX

Explanations

return percentage

The neuron activates on numeric tokens—especially decimal numbers and percentage figures.

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

元々

-1.77

まさか

-1.70

どうやら

-1.67

');

-1.64

これも

-1.63

最初は

-1.61

！」

-1.56

僕も

-1.55

 автомобиль

-1.55

後日

-1.54

POSITIVE LOGITS

of

2.08

 उन

1.66

 bijvoorbeeld

1.59

<strong>

1.56

say

1.53

1.52

ଝ

1.51

༷

1.46

᯾

1.45

етки

1.45

Activations Density 0.011%