Structured Abductive-Deductive-Inductive Reasoning for LLMs via Algebraic Invariants
通过代数不变量实现LLMs的结构归纳-演绎-演绎推理
机构 * DeepThought Solutions(深思科技) ; Department of Computer Science University of Florida(佛罗里达大学计算机科学系)
AI总结 本文提出通过代数不变量实现LLMs的结构归纳-演绎-演绎推理框架,通过五种代数不变量确保推理链中结论的可靠性,防止逻辑不一致的累积。
Comments 10 pages + 3 pages references. Accepted as a poster at the ICLR 2026 Workshop for LLM Reasoning