INDEX

Explanations

years and dates

The neuron primarily activates on four-digit numeric tokens representing years or dates.

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

-"

-1.59

».

-1.58

end

-1.57

;?

-1.56

meture

-1.55

-1.52

 porque

-1.45

-1.42

.'"

-1.42

 perché

-1.38

POSITIVE LOGITS

selben

1.64

1.58

 ucap

1.57

 Hingga

1.48

 saman

1.47

 Kedua

1.46

uyla

1.44

 kenal

1.41

 bluz

1.40

 Menurut

1.38

Activations Density 0.084%