INDEX

Explanations

titles and quoted conditionals

The neuron fires on Latin‐derived or pseudo-Latin stems (e.g. “ped,” “rag,” “velum,” “mo”)—in other words, it spots formal or jargon-style vocabulary of Latin origin.

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

🤸

-1.26

 πρ

-1.23

 größe

-1.23

💃

-1.21

🎊

-1.20

☀️

-1.19

鐿

-1.17

sunny

-1.14

🕺

-1.13

 réguli

-1.09

POSITIVE LOGITS

 иллюстрация

1.33

因为

1.30

Zutaten

1.27

 början

1.24

getSystemService

1.20

 нарушения

1.19

 vielleicht

1.16

 naturen

1.15

 כוס

1.15

 amorosa

1.14

Activations Density 0.000%