{"slug": "bootstrapping-niche-multilingual-code-translation-via-reinforcement-learning", "title": "Bootstrapping Niche Multilingual Code Translation via Reinforcement Learning with Execution-Based Verifiable Supervision", "summary": "Researchers introduced a reinforcement learning pipeline that uses execution-based supervision to improve multilingual code translation across 600 directed language pairs, achieving a 13% average improvement on the new HumanEval-X++ benchmark with a 4B model. The approach, detailed in arXiv:2608.13854v1, expands verifiable Python programs into multilingual pools, trains a reward model on execution outcomes, and optimizes Qwen-3.5 4B and 9B models via GRPO, with the 4B model gaining 21% on mid-tier languages.", "body_md": "arXiv:2608.13854v1 Announce Type: new\nAbstract: Code translation must preserve executable behavior across many programming languages, yet neural code translation has largely focused on a few popular languages such as C++, Java, and Python. This leaves a niche, many-to-many setting where parallel supervision is sparse, producing plausible but non-executable translations. We address this setting with preference-based reinforcement learning driven by execution-based supervision. Our pipeline firstly expands verifiable seed Python programs into a multilingual pool of execution-validated codes. Using the pool, a base LLM generates translation candidates across language pairs, which we label by their execution outcomes. The resulting preferences are used to train a reward model that scores cross-language translation quality. Finally, we optimize our base LLMs with GRPO over 600 directed language pairs (25 x 24) using the reward model as a signal. To evaluate the niche translation capability, we introduce HumanEval-X++, an execution-based benchmark that extends HumanEval-X to a broad many-to-many language space. We evaluate our approach using Qwen-3.5 4B and 9B models. On HumanEval-X++ and existing benchmarks, it yields consistent gains over the untrained baselines. In particular, the 4B model achieves an average improvement of 13% across all languages on HumanEval-X++, with a gain of 21% on mid-tier languages. Our study establishes a reliable approach of data generation, training, and benchmarking, paving the way toward further bootstrapping the quality of many-to-many translation for programming languages.", "url": "https://wpnews.pro/news/bootstrapping-niche-multilingual-code-translation-via-reinforcement-learning", "canonical_source": "https://arxiv.org/abs/2608.13854", "published_at": "2026-08-17 04:00:00+00:00", "updated_at": "2026-08-17 04:14:27.880018+00:00", "lang": "en", "topics": ["machine-learning", "large-language-models", "artificial-intelligence"], "entities": ["arXiv", "HumanEval-X++", "HumanEval-X", "Qwen-3.5", "GRPO"], "alternates": {"html": "https://wpnews.pro/news/bootstrapping-niche-multilingual-code-translation-via-reinforcement-learning", "markdown": "https://wpnews.pro/news/bootstrapping-niche-multilingual-code-translation-via-reinforcement-learning.md", "text": "https://wpnews.pro/news/bootstrapping-niche-multilingual-code-translation-via-reinforcement-learning.txt", "jsonld": "https://wpnews.pro/news/bootstrapping-niche-multilingual-code-translation-via-reinforcement-learning.jsonld"}}