CoMAT: Chain of Mathematically Annotated Thought Improves Mathematical Reasoning

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Leang, Joshua Ong Jun, Gema, Aryo Pradipta, Cohen, Shay B.
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866908767955714048
author Leang, Joshua Ong Jun
Gema, Aryo Pradipta
Cohen, Shay B.
author_facet Leang, Joshua Ong Jun
Gema, Aryo Pradipta
Cohen, Shay B.
contents Mathematical reasoning remains a significant challenge for large language models (LLMs), despite progress in prompting techniques such as Chain-of-Thought (CoT). We present **Chain of Mathematically Annotated Thought (CoMAT)**, which enhances reasoning through two stages: *Symbolic Conversion* (converting natural language queries into symbolic form) and *Reasoning Execution* (deriving answers from symbolic representations). CoMAT operates entirely with a single LLM and without external solvers. Across four LLMs, CoMAT outperforms traditional CoT on six out of seven benchmarks, achieving gains of 4.48% on MMLU-Redux (MATH) and 4.58% on GaoKao MCQ. In addition to improved performance, CoMAT ensures faithfulness and verifiability, offering a transparent reasoning process for complex mathematical tasks
format Preprint
id arxiv_https___arxiv_org_abs_2410_10336
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle CoMAT: Chain of Mathematically Annotated Thought Improves Mathematical Reasoning
Leang, Joshua Ong Jun
Gema, Aryo Pradipta
Cohen, Shay B.
Artificial Intelligence
Computation and Language
Machine Learning
Symbolic Computation
Mathematical reasoning remains a significant challenge for large language models (LLMs), despite progress in prompting techniques such as Chain-of-Thought (CoT). We present **Chain of Mathematically Annotated Thought (CoMAT)**, which enhances reasoning through two stages: *Symbolic Conversion* (converting natural language queries into symbolic form) and *Reasoning Execution* (deriving answers from symbolic representations). CoMAT operates entirely with a single LLM and without external solvers. Across four LLMs, CoMAT outperforms traditional CoT on six out of seven benchmarks, achieving gains of 4.48% on MMLU-Redux (MATH) and 4.58% on GaoKao MCQ. In addition to improved performance, CoMAT ensures faithfulness and verifiability, offering a transparent reasoning process for complex mathematical tasks
title CoMAT: Chain of Mathematically Annotated Thought Improves Mathematical Reasoning
topic Artificial Intelligence
Computation and Language
Machine Learning
Symbolic Computation
url https://arxiv.org/abs/2410.10336