INDEX

Explanations

of them / of which

the neuron detects numeric tokens representing quantities (e.g. counts, measurements, percentages, or other numbers)

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

nº

-1.33

ꙮ

-1.30

 toujours

-1.27

 superbe

-1.27

 normaal

-1.26

 nimmer

-1.26

 exercices

-1.23

 qualquer

-1.23

 magnifique

-1.23

 travail

-1.21

POSITIVE LOGITS

 were

1.50

on

1.42

 that

1.41

 from

1.33

 включая

1.25

 with

1.20

 leichter

1.19

out

1.16

遴

1.16

by

1.13

Activations Density 0.129%