INDEX

Explanations

center, left, align, coordinates

The neuron fires on numeric tokens—especially decimal numbers, measurements, dimensions, and other numeric values in style or data contexts.

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

 because

-2.41

 most

-2.30

.......

-2.23

an

-2.22

........

-2.20

 ....

-2.19

 .....

-2.17

 just

-2.16

so

-2.11

 relatively

-2.09

POSITIVE LOGITS

 offrir

2.14

 envoyer

2.13

 REALLY

2.13

 attirer

2.08

 wondrous

2.06

 harten

2.06

 luscious

2.06

 quintessential

2.06

 sooo

2.03

 avancer

2.03

Activations Density 0.036%