Summary of Enhancing Mathematical Reasoning in Llms with Background Operators, by Jiajun Chen and Yik-cheung Tam
Enhancing Mathematical Reasoning in LLMs with Background Operators
by Jiajun Chen, Yik-Cheung Tam
First submitted to arxiv on: 5 Dec 2024
Categories
- Main: Artificial Intelligence (cs.AI)
- Secondary: None
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed approach uses background operators for mathematical reasoning in large language models (LLMs). The method defines fundamental mathematical predicates as building blocks and develops Prolog solutions for each problem using these predicates. A new corpus, MATH-Prolog, is introduced, derived from the counting and probability categories of the MATH corpus. Self-training with K-fold cross-validation is applied to efficiently augment data. Experimental results show that this approach achieves high accuracy (84.6% on the cross-validated set, 84.8% during fine-tuning) by uncovering new solutions with fully computable inference steps for previously unseen problems. The inclusion of background mathematical predicates in prompts also enhances solution coverage. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Large language models can be used to reason mathematically. To do this, we define basic building blocks and use them to solve math problems. We made a special dataset (MATH-Prolog) by taking some math problems from another dataset (MATH). Then, we tried training the model on itself to make it better at solving math problems. The results show that this approach works well: it found new solutions for math problems it hadn’t seen before and was very accurate. |
Keywords
» Artificial intelligence » Fine tuning » Inference » Probability » Self training
