Search papers, labs, and topics across Lattice.
The paper introduces Framework of Thoughts (FoT), a foundation framework designed to build and optimize dynamic reasoning schemes for large language models, addressing the limitations of static, problem-specific reasoning structures like Chain of Thought. FoT incorporates hyperparameter tuning, prompt optimization, parallel execution, and intelligent caching to improve the performance of reasoning schemes. Experiments implementing Tree of Thoughts, Graph of Thoughts, and ProbTree within FoT demonstrate faster execution, reduced costs, and improved task scores through optimization.
Unleash the hidden potential of LLM reasoning with Framework of Thoughts, a new framework that dynamically optimizes reasoning structures like ToT and GoT for speed, cost, and accuracy.
Prompting schemes such as Chain of Thought, Tree of Thoughts, and Graph of Thoughts can significantly enhance the reasoning capabilities of large language models. However, most existing schemes require users to define static, problem-specific reasoning structures that lack adaptability to dynamic or unseen problem types. Additionally, these schemes are often under-optimized in terms of hyperparameters, prompts, runtime, and prompting cost. To address these limitations, we introduce Framework of Thoughts (FoT)--a general-purpose foundation framework for building and optimizing dynamic reasoning schemes. FoT comes with built-in features for hyperparameter tuning, prompt optimization, parallel execution, and intelligent caching, unlocking the latent performance potential of reasoning schemes. We demonstrate FoT's capabilities by implementing three popular schemes--Tree of Thoughts, Graph of Thoughts, and ProbTree--within FoT. We empirically show that FoT enables significantly faster execution, reduces costs, and achieves better task scores through optimization. We release our codebase to facilitate the development of future dynamic and efficient reasoning schemes.
