![[2602.16512] Framework of Thoughts: A Foundation Framework for Dynamic and Optimized Reasoning based on Chains, Trees, and Graphs](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.16512] Framework of Thoughts: A Foundation Framework for Dynamic and Optimized Reasoning based on Chains, Trees, and Graphs
Summary
The article presents the Framework of Thoughts (FoT), a new foundation framework designed to enhance the reasoning capabilities of large language models through dynamic and optimized prompting schemes.
Why It Matters
As AI continues to evolve, the ability to adapt reasoning structures for diverse problem types is crucial. FoT addresses limitations in existing prompting schemes, offering a more efficient and flexible approach that can significantly improve performance and reduce costs in AI applications.
Key Takeaways
- FoT introduces a general-purpose framework for dynamic reasoning in AI.
- It optimizes existing prompting schemes like Tree of Thoughts and Graph of Thoughts.
- The framework includes features for hyperparameter tuning and intelligent caching.
- Empirical results show FoT enhances execution speed and task performance.
- The codebase is released to support future developments in AI reasoning.
Computer Science > Artificial Intelligence arXiv:2602.16512 (cs) [Submitted on 18 Feb 2026] Title:Framework of Thoughts: A Foundation Framework for Dynamic and Optimized Reasoning based on Chains, Trees, and Graphs Authors:Felix Fricke, Simon Malberg, Georg Groh View a PDF of the paper titled Framework of Thoughts: A Foundation Framework for Dynamic and Optimized Reasoning based on Chains, Trees, and Graphs, by Felix Fricke and 2 other authors View PDF HTML (experimental) Abstract: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, reduc...
