INDEX

Explanations

action or choice

The neuron strongly activates on RPG rule-text tokens—especially those containing numbers or numeric qualifiers (e.g. “5-foot,” “bonus,” “action,” “sphere,” dice-stat entries)—i.e. mechanically quantified terms.

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

 église

-1.45

ּוֹ

-1.39

pullover

-1.39

ﻹ

-1.34

 πρ

-1.26

 schnee

-1.25

 batas

-1.25

satin

-1.24

 wakil

-1.23

giai

-1.20

POSITIVE LOGITS

<em>

1.38

or

1.34

𝑞

1.30

 Janata

1.29

ngua

1.29

ējās

1.29

 стандарт

1.24

 россии

1.23

uring

1.23

he

1.22

Activations Density 0.013%