Large Language Models in Optimization and Decision-Making

The field of optimization and decision-making is witnessing a significant shift towards the integration of large language models (LLMs) to improve the quality and robustness of decision-making processes. This trend is driven by the ability of LLMs to handle uncertainty, provide contextual understanding, and facilitate iterative collaboration between optimization and LLMs agents. The use of LLMs is being explored in various applications, including financial trading, interior design, and mixed-reality window management. Notable papers in this area include DAOpt, which proposes a framework for modeling and evaluating data-driven optimization under uncertainty with LLMs, and SOLID, which integrates mathematical optimization with the contextual capabilities of LLMs. FINRS is also noteworthy for its risk-sensitive trading framework that combines hierarchical market analysis, dual-decision agents, and multi-timescale reward reflection to align trading actions with both return objectives and downside risk constraints. DuoZone is another notable example, which introduces a mixed-initiative XR window management system that combines user-defined spatial layouts with LLM-guided automation.

Sources

DAOpt: Modeling and Evaluation of Data-Driven Optimization under Uncertainty with LLMs

Co-Layout: LLM-driven Co-optimization for Interior Layout

LOBERT: Generative AI Foundation Model for Limit Order Book Messages

FINRS: A Risk-Sensitive Trading Framework for Real Financial Markets

SOLID: a Framework of Synergizing Optimization and LLMs for Intelligent Decision-Making

DuoZone: A User-Centric, LLM-Guided Mixed-Initiative XR Window Management System

Built with on top of