Understanding the Structure of Reasoning in Language Models

Soheil Feizi discusses the structure and dynamics of reasoning in language models, their behavior under uncertainty, and controlling inference-time computation.

Soheil Feizi (University of Maryland, College Park) simons.berkeley.edu/talks/soheil-feizi-university-maryland-c… Trustworthy AI: From Hallucinations to Reliable Autonomy Large language models are becoming increasingly capable reasoners, but reliability requires more than solving individual problems correctly. Models must reason under uncertainty, allocate computation effectively, recognize their limitations, and decide when and how to use external information or tools. In this talk, I will discuss recent work on the structure and dynamics of reasoning in language models, their ability to reason under uncertainty, and mechanisms for controlling inference-time computation. Across these settings, a recurring theme is that strong underlying capabilities do not always translate into reliable reasoning behavior. Understanding this gap—and the internal signals and mechanisms that govern reasoning—will become central to building more reliable and autonomous AI systems.