INDEX

Explanations

numerals and 1 mm/year

The neuron primarily identifies numeric measurement tokens (numbers and values, often with units or decimals).

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

</h4>

-1.52

-1.46

 many

-1.40

ucing

-1.30

たちも

-1.29

，

-1.29

віч

-1.28

ŚĆ

-1.24

↵↵↵↵↵↵↵↵↵↵↵

-1.23

are

-1.23

POSITIVE LOGITS

 показывает

1.59

 отличается

1.49

 представляет

1.44

になっています

1.43

 comenzará

1.41

 видите

1.40

 позволяет

1.32

 зависит

1.31

 Bedürfnisse

1.30

Activations Density 0.250%