OnThePlanning Abilities of OpenAI’s o1 Models: Feasibility, Optimality, and Generalizability

Kevin Wang, Junbo Li, Neel P. Bhatt, Yihan Xi, qiang liu, ufuk topcu, Zhangyang Wang

Keywords: Evaluation, LLM, planning, reasoning, interactive decision-making, Benchmark

TL;DR: We evaluated the GPT4 and o1 on planning tasks, highlighting their strength in problem understanding and identifying challenges in spatial reasoning and generalization.

Abstract: Recent advancements in Large Language Models (LLMs) have showcased their ability to perform complex reasoning tasks, but their effectiveness in planning remains underexplored. In this study, we evaluate the planning capabilities of OpenAI's o1 models across a variety of benchmark tasks, focusing on three key aspects: feasibility, optimality, and generalizability. Through empirical evaluations on constraint-heavy tasks and spatially complex environments, we highlight o1-preview’s strengths in self-evaluation and constraint-following, while also identifying bottlenecks in decision-making and memory management, particularly in tasks requiring robust spatial reasoning. Our results reveal that while o1-preview outperforms GPT-4, the model often generates suboptimal solutions with redundant actions and struggles to generalize effectively in spatially complex tasks. This pilot study provides foundational insights into the planning limitations of LLMs, offering key directions for future research on improving memory management, decision-making, and generalization in LLM-based planning.

Submission Number: 29