INDEX

Explanations

fortified, refractory, Fortis

The neuron detects uses of the root “fortif-,” i.e. terms like “fortified” or “fortification.”

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

al

-2.47

檮

-2.20

赜

-2.19

 öfter

-2.17

ounced

-2.14

 empezaron

-2.13

𝑁

-2.06

ischen

-2.06

 наша

-2.05

 vorsichtig

-2.02

POSITIVE LOGITS

‘

2.77

”

2.30

 alcuni

2.22

jnych

2.16

 یک

2.11

of

2.11

霂

2.11

!”

2.09

”;

2.08

{~

2.05

Activations Density 0.003%