2 papers
cs.LG2026
No Subspace to Track: Non-Identifiability and Optimizer State in Low-Rank Training
Noel Thomas
Memory-efficient optimizers such as GaLore train large language models by projecting gradients onto a rank-r subspace recomputed every T steps, assuming this subspace is a slowly d…
cs.CE2026
BioXArena: Benchmarking LLM Agents on Multi-Modal Biomedical Machine Learning Tasks
Loka Li, Duzhen Zhang, Xingbo Du +11
Large language model (LLM) agents are increasingly capable of automating components of machine learning development, yet existing biomedical benchmarks mainly focus on question ans…