INDEX

Explanations

product features or contents

The neuron is keyed to technical product-specification tokens—especially numbers, measurements, percentages, and performance‐claim 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

 nějak

-1.55

𝑞

-1.45

𝑗

-1.36

说法

-1.31

 бампер

-1.29

我觉得

-1.29

🧫

-1.28

᾽

-1.26

我现在

-1.23

ὠ

-1.22

POSITIVE LOGITS

 kapsul

1.46

 rôle

1.40

 sirop

1.38

 even

1.37

 soigneusement

1.35

kym

1.34

 sains

1.34

 your

1.34

 sogar

1.33

 sirup

1.32

Activations Density 0.076%