Challenge: Large language models (LLMs) have shown excellent mastering of human language but struggle in real-world applications that require mathematical problem-solving.
Approach: They propose a pipeline to train a general Math-Critique model from the LLM itself to provide feedback signals and employ rejective fine-tuning and direct preference optimization over the Llm's own generations for data collection.
Outcome: The proposed pipeline outperforms existing LLMs that could be two times larger.

Similar Papers

MARIO: MAth Reasoning with code Interpreter Output - A Reproducible Pipeline (2024.findings-acl)

Copied to clipboard

Challenge: Large language models lack mathematical reasoning, a hurdle on the path to true artificial general intelligence.
Approach: They propose a protocol for fine-tuning large language models with a Python code interpreter to enhance the text analysis of the LLMs.
Outcome: The proposed protocol improves the performance of a 7B-parameter LLM on the GSM8K and MATH datasets while allowing for an outlier-free value model-based inference method.
Large Language Models for Mathematical Reasoning: Progresses and Challenges (2024.eacl-srw)

Copied to clipboard

Challenge: a survey examines the landscape of mathematical problem-solving techniques . large language models have proven to be potent assets in unraveling nuances of mathematical reasoning .
Approach: They examine the evolution of Large Language Models (LLMs) for solving mathematical problems . they examine the spectrum of LLM-oriented techniques proposed for solving math problems - and their challenges .
Outcome: The survey examines the spectrum of proposed LLM-oriented techniques in solving math problems.
Self-Explore: Enhancing Mathematical Reasoning in Language Models with Fine-grained Rewards (2024.findings-emnlp)

Copied to clipboard

Challenge: Recent studies have shown that large language models can solve complex reasoning tasks with Chain-of-Thought Prompting.
Approach: They propose a training method where the LLM is tasked to explore the first wrong step within the rationale and use such signals as fine-grained rewards for further improvement.
Outcome: The proposed model improves on the GSM8K and MATH test sets by 11.57% and 2.89% on average compared to supervised fine-tuning (SFT).
Breaking Language Barriers in Multilingual Mathematical Reasoning: Insights and Observations (2024.findings-emnlp)

Copied to clipboard

Challenge: Existing research focuses on developing powerful large language models for mathematical reasoning within monolingual languages.
Approach: They propose to use translation to build powerful multilingual math reasoning models . they propose different training strategies to build xMR LLMs that outperform open-source LLM .
Outcome: The proposed model outperforms open-source LLMs and surpasses ChatGPT in few-shot scenarios.
MathFusion: Enhancing Mathematical Problem-solving of LLM through Instruction Fusion (2025.acl-long)

Copied to clipboard

Challenge: Large Language Models (LLMs) have shown impressive progress in mathematical problem-solving . current approaches to enhance mathematical reasoning focus on instance-level modifications .
Approach: They propose a framework that enhances mathematical reasoning through cross-problem instruction synthesis.
Outcome: The proposed framework boosts mathematical reasoning by 18.0 points while maintaining high data efficiency.
Three Questions Concerning the Use of Large Language Models to Facilitate Mathematics Learning (2023.findings-emnlp)

Copied to clipboard

Challenge: After the pandemic, e-learning has become part of mainstream education.
Approach: They propose to integrate large language models (LLMs) into educational settings to enhance students' mathematical problem-solving skills by providing adaptive feedback.
Outcome: The proposed model can generate free-text rationalizations and misinterpret meanings and can also misinterprét students' answers.
More Data or Better Data? A Critical Analysis of Data Selection and Synthesis for Mathematical Reasoning (2025.emnlp-industry)

Copied to clipboard

Challenge: Despite various proposed data construction methods, their practical utility in real-world pipelines remains underexplored.
Approach: They conduct a comprehensive analysis of open-source datasets and data synthesis techniques for mathematical reasoning under a unified pipeline designed to mirror training and deployment scenarios.
Outcome: The proposed pipelines mirror training and deployment scenarios and are suitable for industrial applications.
From A and B to A+B: Can Large Language Models Solve Compositional Math Problems? (2025.emnlp-main)

Copied to clipboard

Challenge: Existing studies that create problem variants by adding perturbations to a single problem focus on the interaction between problems.
Approach: They propose a pipeline with 98.2% accuracy to combine two original problems with a logical connection and to evaluate LLMs' generalization ability on the compositional problems.
Outcome: The proposed pipeline can combine two original problems with a logical connection to get a new math problem and evaluate its compositional generalization on the compositional problems.
Teaching Language Models to Self-Improve through Interactive Demonstrations (2024.naacl-long)

Copied to clipboard

Challenge: Large language models (LLMs) have been shown to improve performance on downstream tasks by prompting them to analyze and revise their outputs.
Approach: They propose a training algorithm that prompts large language models to analyze and revise their own outputs and uses this feedback to train the small model.
Outcome: The proposed approach improves LLaMA-7B's performance on math and reasoning tasks by up to 7.13%.
CriticBench: Benchmarking LLMs for Critique-Correct Reasoning (2024.findings-acl)

Copied to clipboard

Challenge: CriticBench is a benchmark designed to assess LLMs’ abilities to critique and refine their reasoning across a variety of tasks.
Approach: They propose a benchmark to assess LLMs' ability to critique and correct reasoning across a variety of tasks.
Outcome: The proposed benchmark examines the performance of 17 large language models in generation, critique, and correction reasoning.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations