INDEX

Explanations

events and their status

The neuron is selectively activating on numeric tokens—especially years, dates, section or citation numbers, and other digit-based references in legal or official text.

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

-1.46

 might

-1.40

이나

-1.34

 probably

-1.32

 better

-1.27

↵↵↵

-1.27

 definitely

-1.26

 like

-1.25

if

-1.23

 especially

-1.23

POSITIVE LOGITS

ּוֹ

2.05

ֶּ

1.73

ִּ

1.72

ַּ

1.68

וּ

1.52

 pursuant

1.40

ְּ

1.39

 boîtier

1.38

 bénéfice

1.31

进行了

1.30

Activations Density 0.038%