AI RESEARCH
Your Model Diversity, Not Method, Determines Reasoning Strategy
arXiv CS.AI
•
ArXi:2604.10827v1 Announce Type: new Compute scaling for LLM reasoning requires allocating budget between exploring solution approaches ($breadth$) and refining promising solutions ($depth$). Most methods implicitly trade off one for the other, yet why a given trade-off works remains unclear, and validation on a single model obscures the role of the model itself.