INDEX

Explanations

academic concepts

This neuron spots formal academic or scientific nominalizations (especially words ending in “-ion,” like regulation, generation, process, description, reaction, perception) that serve as key topic terms in scholarly writing.

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

but

-1.53

 give

-1.30

 phenomenal

-1.20

 когда

-1.17

 huge

-1.16

 enormous

-1.15

 incredible

-1.13

 gives

-1.13

 because

-1.11

 good

-1.08

POSITIVE LOGITS

of

1.77

alors

1.33

を

1.15

かというと

1.14

が

1.13

ximos

1.09

 activer

1.09

yng

1.09

และ

1.07

IDAS

1.07

Activations Density 0.116%