INDEX

Explanations

Legendre, harmonic, polynomials, skater

The neuron is tuned to spot technical mathematical terms—especially names of functions or polynomial‐type objects (e.g. “harmonic,” “polynomials,” “wavelets,” “coefficients,” etc.).

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

 Finally

-1.27

ǉ

-1.27

就没

-1.22

 levou

-1.22

佷

-1.21

way

-1.20

is

-1.20

 itself

-1.19

 entrou

-1.19

 الصيف

-1.18

POSITIVE LOGITS

 segítség

1.27

are

1.27

 hjälpa

1.22

 pirm

1.21

过敏

1.20

 papild

1.20

 هستند

1.19

our

1.18

sont

1.17

any

1.16

Activations Density 0.033%