SMaRT: Select, Mix, and ReinvenT - A Strategy Fusion Framework for LLM-Driven Reasoning and Planning

2025-10-20

AI Summary

Existing large language models (LLMs) rely on single, fixed reasoning strategies for complex task automation, limiting their generalization and robustness. This paper introduces SMaRT, a framework that enables adaptive decision-making and planning by dynamically selecting, mixing, and recombining multiple reasoning strategies. Its core innovation is treating the LLM as an intelligent fusion engine—rather than a static evaluator—for the first time enabling cross-strategy collaborative reasoning and feedback-driven online calibration. SMaRT integrates prompt engineering, strategy selection algorithms, and dynamic recombination mechanisms. Evaluated on reasoning, planning, and sequential decision-making benchmarks, it achieves significant improvements in solution quality, constraint adherence, and overall performance. These results empirically validate the critical role of multi-strategy collaboration in enhancing LLM robustness and reliability.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Safety and Robustness
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Large Language Models (LLMs) have redefined complex task automation with exceptional generalization capabilities. Despite these advancements, state-of-the-art methods rely on single-strategy prompting, missing the synergy of diverse reasoning approaches. No single strategy excels universally, highlighting the need for frameworks that fuse strategies to maximize performance and ensure robustness. We introduce the Select, Mix, and ReinvenT (SMaRT) framework, an innovative strategy fusion approach designed to overcome this constraint by creating balanced and efficient solutions through the seamless integration of diverse reasoning strategies. Unlike existing methods, which employ LLMs merely as evaluators, SMaRT uses them as intelligent integrators, unlocking the "best of all worlds" across tasks. Extensive empirical evaluations across benchmarks in reasoning, planning, and sequential decision-making highlight the robustness and adaptability of SMaRT. The framework consistently outperforms state-of-the-art baselines in solution quality, constraint adherence, and performance metrics. This work redefines LLM-driven decision-making by pioneering a new paradigm in cross-strategy calibration, unlocking superior outcomes for reasoning systems and advancing the boundaries of self-refining methodologies.

Problem

• Fusing diverse reasoning strategies to maximize LLM performance
• Overcoming limitations of single-strategy prompting in complex tasks
• Enhancing robustness through intelligent integration of multiple approaches

Innovation

• Fuses diverse reasoning strategies for balanced solutions
• Uses LLMs as intelligent integrators of multiple approaches
• Pioneers cross-strategy calibration for superior decision-making