Mengzhang Cai — accepted AI-conference papers

2025

MetaLadder: Ascending Mathematical Solution Quality via Analogical-Problem Reasoning Transfer

Honglin Lin, Zhuoshi Pan, Qizhi Pei, Xin Gao, Yu Li, Mengzhang Cai +2

EMNLP 2025

Large Language Models (LLMs) have demonstrated promising capabilities in solving mathematical reasoning tasks, leveraging Chain-of-Thought (CoT) data as a vital component in guiding answer generation. Current paradigms typically generate CoT and answers directly for a given problem, diverging from h

Cited by 0 Source ↗PDF ↗code ↗Scholar ↗

2025

Middo: Model-Informed Dynamic Data Optimization for Enhanced LLM Fine-Tuning via Closed-Loop Learning

Zinan Tang, Xin Gao, Qizhi Pei, Zhuoshi Pan, Mengzhang Cai, Jiang Wu +2

EMNLP 2025

Supervised Fine-Tuning (SFT) Large Language Models (LLM) fundamentally rely on high-quality training data. While data selection and data synthesis are two common strategies to improve data quality, existing approaches often face limitations in static dataset curation that fail to adapt to evolving m

Cited by 0 Source ↗PDF ↗code ↗Scholar ↗