INDEX

Explanations

court, earth, faces, me, year

The neuron picks up on words that signal spatial relationships or movement—prepositions (in, on, over, within, down, at, for), motion verbs (approached, walked), and other location‐ or direction‐bearing 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

one

-1.09

 young

-1.02

you

-1.01

 [''],

-0.99

⅕

-0.98

got

-0.95

⅔

-0.95

One

-0.94

To

-0.94

～。

-0.94

POSITIVE LOGITS

苎

1.26

ませんでした

1.24

 zwarte

1.21

瘿

1.19

despite

1.19

 avaliação

1.16

 zwart

1.16

ijs

1.14

diep

1.14

 overflowing

1.13

Activations Density 0.062%