Learning to Solve Complex Problems via Dataset Decomposition
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866914346360111104 |
|---|---|
| author | Zhao, Wanru Caccia, Lucas Shi, Zhengyan Kim, Minseon Xu, Weijia Sordoni, Alessandro |
| author_facet | Zhao, Wanru Caccia, Lucas Shi, Zhengyan Kim, Minseon Xu, Weijia Sordoni, Alessandro |
| contents | Curriculum learning is a class of training strategies that organizes the data being exposed to a model by difficulty, gradually from simpler to more complex examples. This research explores a reverse curriculum generation approach that recursively decomposes complex datasets into simpler, more learnable components. We propose a teacher-student framework where the teacher is equipped with the ability to reason step-by-step, which is used to recursively generate easier versions of examples, enabling the student model to progressively master difficult tasks. We propose a novel scoring system to measure data difficulty based on its structural complexity and conceptual depth, allowing curriculum construction over decomposed data. Experiments on math datasets (MATH and AIME) and code generation datasets demonstrate that models trained with curricula generated by our approach exhibit superior performance compared to standard training on original datasets. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_20296 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Learning to Solve Complex Problems via Dataset Decomposition Zhao, Wanru Caccia, Lucas Shi, Zhengyan Kim, Minseon Xu, Weijia Sordoni, Alessandro Machine Learning Curriculum learning is a class of training strategies that organizes the data being exposed to a model by difficulty, gradually from simpler to more complex examples. This research explores a reverse curriculum generation approach that recursively decomposes complex datasets into simpler, more learnable components. We propose a teacher-student framework where the teacher is equipped with the ability to reason step-by-step, which is used to recursively generate easier versions of examples, enabling the student model to progressively master difficult tasks. We propose a novel scoring system to measure data difficulty based on its structural complexity and conceptual depth, allowing curriculum construction over decomposed data. Experiments on math datasets (MATH and AIME) and code generation datasets demonstrate that models trained with curricula generated by our approach exhibit superior performance compared to standard training on original datasets. |
| title | Learning to Solve Complex Problems via Dataset Decomposition |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2602.20296 |