SAAS: Solving Ability Amplification Strategy for Enhanced Mathematical Reasoning in Large Language Models
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866914962166775808 |
|---|---|
| author | Kim, Hyeonwoo Gim, Gyoungjin Kim, Yungi Kim, Jihoo Kim, Byungju Lee, Wonseok Park, Chanjun |
| author_facet | Kim, Hyeonwoo Gim, Gyoungjin Kim, Yungi Kim, Jihoo Kim, Byungju Lee, Wonseok Park, Chanjun |
| contents | This study presents a novel learning approach designed to enhance both mathematical reasoning and problem-solving abilities of Large Language Models (LLMs). We focus on integrating the Chain-of-Thought (CoT) and the Program-of-Thought (PoT) learning, hypothesizing that prioritizing the learning of mathematical reasoning ability is helpful for the amplification of problem-solving ability. Thus, the initial learning with CoT is essential for solving challenging mathematical problems. To this end, we propose a sequential learning approach, named SAAS (Solving Ability Amplification Strategy), which strategically transitions from CoT learning to PoT learning. Our empirical study, involving an extensive performance comparison using several benchmarks, demonstrates that our SAAS achieves state-of-the-art (SOTA) performance. The results underscore the effectiveness of our sequential learning approach, marking a significant advancement in the field of mathematical reasoning in LLMs. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2404_03887 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | SAAS: Solving Ability Amplification Strategy for Enhanced Mathematical Reasoning in Large Language Models Kim, Hyeonwoo Gim, Gyoungjin Kim, Yungi Kim, Jihoo Kim, Byungju Lee, Wonseok Park, Chanjun Computation and Language Artificial Intelligence This study presents a novel learning approach designed to enhance both mathematical reasoning and problem-solving abilities of Large Language Models (LLMs). We focus on integrating the Chain-of-Thought (CoT) and the Program-of-Thought (PoT) learning, hypothesizing that prioritizing the learning of mathematical reasoning ability is helpful for the amplification of problem-solving ability. Thus, the initial learning with CoT is essential for solving challenging mathematical problems. To this end, we propose a sequential learning approach, named SAAS (Solving Ability Amplification Strategy), which strategically transitions from CoT learning to PoT learning. Our empirical study, involving an extensive performance comparison using several benchmarks, demonstrates that our SAAS achieves state-of-the-art (SOTA) performance. The results underscore the effectiveness of our sequential learning approach, marking a significant advancement in the field of mathematical reasoning in LLMs. |
| title | SAAS: Solving Ability Amplification Strategy for Enhanced Mathematical Reasoning in Large Language Models |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2404.03887 |