INDEX

Explanations

game entities and jargon

The neuron fires on capitalized or domain‐specific names, acronyms, and jargon—e.g. game titles, character or boss names, skill/stat abbreviations, and other specialized 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

 cadeira

-1.33

Mutability

-1.26

 хорошая

-1.18

Lister

-1.16

 оно

-1.15

 vagyok

-1.13

chalten

-1.13

listbox

-1.12

 Capitán

-1.12

 Dominicana

-1.12

POSITIVE LOGITS

On

1.28

"}";

1.24

1.23

 weiterhin

1.17

However

1.17

 caráter

1.16

?,?,

1.14

 kemudian

1.11

−/−

1.11

Activations Density 0.011%