INDEX

Explanations

years or centuries starting with 19

This neuron detects numeric date expressions, especially years and centuries.

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

 then

-1.37

 muelle

-1.37

 there

-1.34

 Ayuntamiento

-1.30

 thier

-1.30

 genellikle

-1.27

ὺς

-1.26

 wspar

-1.23

 xám

-1.20

Nguy

-1.20

POSITIVE LOGITS

ging

1.37

tacular

1.33

legg

1.23

makeText

1.13

です

1.13

skap

1.13

("<

1.11

lless

1.11

alanta

1.11

----

1.10

Activations Density 0.193%