INDEX

Explanations

period followed by sentence starters

The neuron activates on numeric tokens that include decimal points (e.g. “33.5,” “22.875,” “97,” etc.).

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

３

-1.84

６

-1.82

５

-1.77

 veliko

-1.72

４

-1.71

-1.69

７

-1.67

Τι

-1.64

 Meningkatkan

-1.60

Οι

-1.59

POSITIVE LOGITS

かもし

1.59

 maniere

1.54

 ponga

1.48

 repens

1.47

と言って

1.46

is

1.42

 laurent

1.35

我们就

1.34

 anything

1.34

 élev

1.33

Activations Density 0.198%