Holistic Data Scheduler for LLM Pre-training via Multi-Objective Reinforcement Learning Paper • 2606.24133 • Published 16 days ago • 11
CausalMix: Data Mixture as Causal Inference for Language Model Training Paper • 2607.01104 • Published 8 days ago • 19
