INDEX

Explanations

The neuron specializes in spotting numeric tokens (including integers, decimals, and numbers embedded in words).

list of categories or types

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Negative Logits

 this

-1.70

in

-1.63

ditionally

-1.60

 only

-1.51

complished

-1.45

 when

-1.42

 either

-1.41

one

-1.40

at

-1.40

 which

-1.38

POSITIVE LOGITS

etc

2.69

などの

1.92

<bos>

1.75

などを

1.74

etc

1.74

 otros

1.66

 diversos

1.66

 wielu

1.59

などが

1.58

その他

1.57

Activations Density 0.009%