INDEX

Explanations

19th century years

The neuron strongly activates on numeric tokens—particularly years, centuries, and other date-related 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

 wichtigste

-1.26

basically

-1.26

沵

-1.25

essentially

-1.23

usions

-1.23

ergies

-1.23

became

-1.22

bücher

-1.22

 hébergement

-1.22

 opulent

-1.21

POSITIVE LOGITS

 when

1.52

If

1.41

if

1.36

and

1.30

 people

1.20

 When

1.19

 choosing

1.16

 what

1.15

Let

1.12

 There

1.11

Activations Density 0.064%