kalomaze's kalomazing blog

/ let's all love learning /

/ let's all love learning /

/ let's all love learning /

kalomaze


i figured a dedicated instance where i can share longform explanations of the projects i'm working on outside of my twitter (https://x.com/kalomaze) would be appreciated, so here we are.

my current focus is on improving generalized preference modeling. i'll be sharing my findings and empirically noting any challenges i face from here on out, as well as experimenting with novel online RL rewards through GRPO.

you can find my WIP efforts that i've published so far under this org:

https://huggingface.co/Quest-AI

more detailed writing to come soon!


\ let's all love learning \

\ let's all love learning \

\ let's all love learning \