Reposted by Nicholas Robertsalbertge.bsky.social @albertge.bsky.social · 08/05/2025Online data mixing reduces training costs for foundation models, but faces challenges: ⚠️ Human-defined domains miss semantic nuances ⚠️ Limited eval accessibility ⚠️ Poor scalability Introducing 🎵R&B: first regroup data, then dynamically reweight domains during training! 153
Reposted by Nicholas RobertsFred Sala @fredsala.bsky.social · 23/04/2025Today at @iclr-conf.bsky.social, come chat with @changho.bsky.social about what types of data drive weak-to-strong generalization! 0103
Reposted by Nicholas RobertsFred Sala @fredsala.bsky.social · 11/12/2024First up at #NeurIPS2024 from our group, our work on labeling via programmatic distillation (a spotlight!). Label your data orders of magnitude faster and cheaper — come join us today at Poster Session 2 East for a demo! 0158