excited to see frens at #icml2026 & present
🐟 Olmix: efficient data mixing under token constraints & evolving data domains 🐡 How2Everything: mining the web for diverse procedural tasks for train & eval
🐠 happy to chat data & evals, both pre & post-training
Users praise Kyle Lo's GIFs during his Olmix and How2Everything research presentation at ICML 2026 as impressively creative.
No Digg Deeper questions have been answered for this story yet.
more on Olmix from @MayeeChen
extended paper: https://arxiv.org/abs/2602.12237
Data mixing - determining ratios across your training datasets - matters a lot for model quality. While building Olmo 3, we learned it’s hard to set up a method that finds a strong mix, and hard to maintain that mix as datasets change throughout development. Introducing Olmix👇
more on How2Everything from @YapeiChang
extended paper: https://arxiv.org/abs/2602.08808v1
How2Everything will appear in ICML 2026! See you in Korea 🫡
We mine the web's procedural knowledge to better evaluate & train LLMs to generate valid step-by-step instructions, read more at: 🔗 https://arxiv.org/pdf/2602.08808
@kylelostat gif game is crazy
