INDEX

Explanations

in the

The neuron activates primarily on numeric tokens (especially scores, statistics, and other decimal-number measurements).

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

 Jika

-1.73

 اگر

-1.70

 यदि

-1.63

 lèvres

-1.63

 Если

-1.60

 ถ้า

-1.60

 Nếu

-1.59

 verlie

-1.57

Kích

-1.55

 Năm

-1.55

POSITIVE LOGITS

 області

1.42

1.33

 modes

1.31

熒

1.26

=\

1.26

 also

1.25

せた

1.25

 were

1.25

illem

1.24

Activations Density 0.024%