Dr. Boot: Bootstrapping Program Synthesis Language Models to Perform Repairing

2025-07-20 · Noah van der Vleuten arxiv

Language models for program synthesis are usually trained and evaluated on programming competition datasets (MBPP, APPS). However, these datasets are limited in size and quality, while these language models are extremely data hungry. Additionally, the language models have a misaligned program synthesis process compared to humans. While humans iteratively develop code with the help of a compiler, most program synthesis models currently produce code in one go. To solve these issues, we introduce a bootstrapping algorithm for program synthesis, that supports teaching models how to repair. We show that bootstrapping consistently outperforms regular fine-tuning. Compared to other work, our bootstrapped model performs on par with fine-tuned models that are 68\% larger. Notably, bootstrapping with repairing also improves non-repairing performance compared to regular bootstrapping during inference. However, on our models, repairing during inference is likely inferior to simply sampling the same number of solutions. Furthermore, we find that there are issues with the example test cases in the training portion of the APPS dataset that are valuable to the community, as many repairing and reinforcement learning methods rely on them.

📄 PDF Abstract BibTeX arXiv:2507.15889

Code (0)

등록된 구현이 없습니다.

Tasks

Reinforcement Learning Program Synthesis

Bootstrapping Code Translation with Weighted Multilanguage Exploration

2026-01-07 · Yuhan Wu, Huan Zhang, Wei Cheng, Chen Shen 외 arxiv

Code translation across multiple programming languages is essential yet challenging due to two vital obstacles: scarcity of parallel data paired with executable test oracles, and optimization imbalance when handling dive…

Reinforcement Learning Code Translation

Stabilizing Off-Policy Q-Learning via Bootstrapping Error Reduction

2019-06-03 · NeurIPS 2019 12 · Aviral Kumar, Justin Fu, George Tucker, Sergey Levine

Off-policy reinforcement learning aims to leverage experience collected from prior policies for sample-efficient learning. However, in practice, commonly used off-policy approximate dynamic programming methods based on Q…

continuous-control Continuous Control Q-Learning Reinforcement Learning

Relative Entropy Regularized Reinforcement Learning for Efficient Encrypted Policy Synthesis

2025-06-14 · Jihoon Suh, Yeongjun Jang, Kaoru Teranishi, Takashi Tanaka

We propose an efficient encrypted policy synthesis to develop privacy-preserving model-based reinforcement learning. We first demonstrate that the relative-entropy-regularized reinforcement learning framework offers a co…

Model-based Reinforcement Learning Privacy Preserving Quantization reinforcement-learning+1

Bootstrapping NLP tools across low-resourced African languages: an overview and prospects

2022-10-21 · C. Maria Keet

Computing and Internet access are substantially growing markets in Southern Africa, which brings with it increasing demands for local content and tools in indigenous African languages. Since most of those languages are l…

Diversity Text Generation

Bootstrap Your Own Context Length

2024-12-25 · Liang Wang, Nan Yang, Xingxing Zhang, Xiaolong Huang 외

We introduce a bootstrapping approach to train long-context language models by exploiting their short-context capabilities only. Our method utilizes a simple agent workflow to synthesize diverse long-context instruction …

Language Modeling Language Modelling