Re:Explorative Modeling – A new axis for pre-training
by albertomr3 on Friday July 31, 2026 @11:56PM
(#49130885)
…does K also have to scale? I’d care a lot about wall-clock and peak memory, especially for local/on-device inference.
[ Parent
| on HN ]
Re:Explorative Modeling – A new axis for pre-training
by alexig on Saturday August 01, 2026 @01:40AM
(#49131380)
you can scale it based on the number of modes in your distribution, and keep it low when its not a big issue (e.g. 2/3 worked pretty well). also, this doesn't affect inference at all so on device can still be fast.