CSE DSI Machine Learning Seminar with Yuandong Tian (Meta AI Research)

Towards a Unified Framework of Neural and Symbolic Decision Making

Large Language Models (LLMs) have made impressive achievement while still struggling with complex tasks that require advanced reasoning, planning, and optimization, which demand a deeper level of thinking than simple chains of thought. Conversely, traditional symbolic solvers provide precise, guaranteed solutions to well-defined problems, but they lack the flexibility to handle more general problems described in natural language. This talk explores unified frameworks that integrate neural and symbolic components, leveraging the strengths of both. We will discuss pure neural models that benefit from symbolic outputs, hybrid and end-to-end systems, and a recent discovery showing that gradient descent solution in some specific reasoning tasks can be completely explained by advanced algebraic (and thus symbolic) objects, such as groups and semi-rings, showing possibility for a deeper unification of these paradigms from first principles.

Yuandong Tian is a Research Scientist Director in Meta AI Research (FAIR), leading the group of reasoning, planning and decision-making with Large Language Models (LLMs). He is the project lead for OpenGo project that beats professional players with a single GPU during inference, serves as the main mentor of StreamingLLM and GaLore that improve the training and inference of LLM, and is the first-author recipient of 2021 ICML Outstanding Paper Honorable Mentions and 2013 ICCV Marr Prize Honorable Mentions, and also received the 2022 CGO Distinguished Paper Award. Prior to that, he worked in Google Self-driving Car team in 2013-2014 and received a Ph.D in Robotics Institute, Carnegie Mellon University in 2013. He has been appointed as area chairs for NeurIPS, ICML, AAAI, CVPR and AIStats.