dynamic

This paper introduces a framework to convert static tasks into dynamic multi-turn conversations to evaluate how well LLMs track evolving user intent, finding that strong static performance does not transfer to dynamic settings.

This paper introduces DynMuon, a dynamic spectral shaping optimizer that schedules the update parameter p from positive to mildly negative during training, consistently achieving lower validation loss and requiring 10.6-26.5% fewer steps than the standard Muon optimizer.