>>109481623
>or is it just too much of a hassle to bother with?
Yeah. see the excerpt below:
=== external wiki: prime patriarchy ===
after years of mystery and mythologizing on men as demi-god figures as their sexual desires bloom, an appropriately-aged boy is introduced to the girls' boarding school and embedded for a week or more. the girls stand in awe as an object of legend appears before them, the boy learns to assert his dominion while exploring tastes and preferences, and they all get to act on their every hormonal urge.
=== external wiki: prime patriarchy ===
the boy discovers his dominion, the girls discovers her subservience, and both pair their revelation with the loss of their virginities. this is the crux of the visitation’s formative power. that week of almost nonstop sex, for both instruction and pleasure, teaches young men the lustful animal nature of women that must be controlled and tamed, as well as the dangerous consequences of greedily surrounding oneself with *too many* sex-crazed slaves. being overwhelmed and brought to the brink of exhaustion is by design.
Splitting them into smaller chunks might result in incoherence. Soo, the option is to configure the frontend (for example, via ST) to use larger chunks. But, doing so means that within the same document, semantically independent sentences will be mixed together. This results in inefficiency and reduced accuracy, because the embed model might retrieve a chunk containing sentences with points X, Y, and Z for query about X which technically mean token bloat, context poisoning, moar shekels to Dario, Sammy, or President Xi.
