LLMs for Mathematical Reasoning
First page
Paper summary
A survey of the fast-growing literature on using LLMs for mathematical reasoning, from arithmetic word problems to theorem proving.
Ask this paper
01
Task landscape: Covers math word problems, formal theorem proving, geometry, and scientific reasoning, showing how each sub-area stresses different LLM capabilities.
02
Methods inventory: Catalogs chain-of-thought, program-aided, tool-using, self-consistency, and verifier-based approaches with benchmark numbers for each.
03
Data and evaluation: Maps the key training and evaluation datasets (GSM8K, MATH, MiniF2F, etc.) and discusses evaluation pitfalls like contamination.
04
Directions: Highlights open problems such as robust multi-step reasoning, integration with formal verifiers, and bridging informal and formal math.