Combinatorial Synthesis: Scaling Code RLVR via Atomic Decomposition and Recombination

• URL: http://arxiv.org/abs/2605.31058v1
• Date: Fri, 29 May 2026 09:29:32 GMT
• Title: Combinatorial Synthesis: Scaling Code RLVR via Atomic Decomposition and Recombination
• Abstract summary: Reinforcement Learning with Verifiable Rewards (RLVR) has emerged as the cornerstone for shaping the remarkable coding abilities of Large Language Models (LLMs)<n>We propose Atomic Decomposition and Recombination (ADR), a novel framework that generates verifiable code tasks via decomposition into atomic elements and controlled recombination.<n>Our work sheds light on a new paradigm for novel code task synthesis and scalable RLVR training.
• Score: 70.27578249309103
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Reinforcement Learning with Verifiable Rewards (RLVR) has recently emerged as the cornerstone for shaping the remarkable coding abilities of Large Language Models (LLMs). However, the scalability of RLVR is severely constrained by the scarcity of sufficiently challenging verifiable code tasks that target near the model's edge of competence. Prior studies often rely on heuristic seed expansions for data synthesis, which severely limits both novelty and difficulty. Consequently, the training value of such data fails to scale proportionally with the size of its synthesis. To this end, we propose Atomic Decomposition and Recombination (ADR), a novel framework that generates verifiable code tasks via decomposition into atomic elements and controlled recombination, thereby enabling the generation of genuinely novel and challenging verifiable code tasks. Experiments and analysis demonstrate that ADR achieves superior originality, difficulty, diversity, and test quality over existing baselines, and consistently delivers greater improvements in code ability across RLVR in diverse downstream domains, including algorithmic programming, tool usage, and data science. Our work sheds light on a new paradigm for novel code task synthesis and scalable RLVR training.

Related papers

• Evolutionary Task Discovery: Advancing Reasoning Frontiers via Skill Composition and Complexity Scaling [10.652489725330401]
  Evoutionary Task Discovery (EvoTD) is a framework that treats data as a directed search over a dual-axis manifold of Algorithmic Skills and Complexity Attributes.<n>We introduce structured evolutionary operators to navigate this space: a Crossover operator that synthesizes novel skill compositions to enhance diversity, and a Parametric Mutation operator that scales structural constraints.<n>EvoTD delivers substantial reasoning gains that generalize consistently across model architectures, pretraining regimes, and scales.
  arXiv  Detail & Related papers  (2026-05-12T07:26:40Z)
• Understanding by Reconstruction: Reversing the Software Development Process for LLM Pretraining [66.89012795621349]
  Large Language Models (LLMs) often struggle with the deep, long-horizon reasoning required for complex software engineering.<n>We propose a novel paradigm: understanding via reconstruction.<n>We introduce a framework that synthesizes latent agentic trajectories using a multi-agent simulation.
  arXiv  Detail & Related papers  (2026-03-11T09:23:20Z)
• Beyond Quantity: Trajectory Diversity Scaling for Code Agents [51.71414642763219]
  Trajectory Diversity Scaling is a data synthesis framework for code agents that scales performance through diversity rather than raw volume.<n> TDScaling integrates four innovations: (1) a Business Cluster mechanism that captures real-service logical dependencies; (2) a blueprint-driven multi-agent paradigm that enforces trajectory coherence; and (3) an adaptive evolution mechanism that steers toward long-tail scenarios.
  arXiv  Detail & Related papers  (2026-02-03T07:43:03Z)
• Golden Goose: A Simple Trick to Synthesize Unlimited RLVR Tasks from Unverifiable Internet Text [90.21130870640057]
  Golden Goose is a trick to synthesize unlimited RLVR tasks from unverifiable internet text.<n>We leverage reasoning-rich unverifiable corpora to synthesize GooseReason-0.7M, a large-scale RLVR dataset.<n>We also deploy Golden Goose in a real-world setting, synthesizing RLVR tasks from raw FineWeb scrapes for the cybersecurity domain.
  arXiv  Detail & Related papers  (2026-01-30T13:39:11Z)
• X-Coder: Advancing Competitive Programming with Fully Synthetic Tasks, Solutions, and Tests [47.271827881215295]
  We propose a fully synthetic approach to training Code LLMs with entirely generated tasks, solutions, and test cases.<n>To support this, we leverage feature-based synthesis to propose a novel data synthesis pipeline called SynthSmith.<n> SynthSmith shows strong potential in producing diverse and challenging tasks, along with verified solutions and tests.<n>X-Coder model series achieves a notable pass rate of 62.9 avg@8 on LiveCodeBench v5 and 55.8 on v6, outperforming DeepCoder-14B-Preview and AReal-boba2-14B.
  arXiv  Detail & Related papers  (2026-01-11T15:22:33Z)
• EvoSyn: Generalizable Evolutionary Data Synthesis for Verifiable Learning [63.03672166010434]
  We introduce an evolutionary, task-agnostic, strategy-guided, executably-checkable data synthesis framework.<n>It jointly synthesizes problems, diverse candidate solutions, and verification artifacts.<n>It iteratively discovers strategies via a consistency-based evaluator that enforces agreement between human-annotated and strategy-induced checks.
  arXiv  Detail & Related papers  (2025-10-20T11:56:35Z)
• VERIRL: Boosting the LLM-based Verilog Code Generation via Reinforcement Learning [32.974199255760944]
  We introduce a reinforcement learning framework tailored for Verilog code generation.<n>To tackle the problem of sparse and noisy reward signals, we propose a Trace-back based Rescore mechanism.<n>To mitigate catastrophic forgetting and overfitting during RL fine-tuning, we introduce a sample-balanced weighting strategy.
  arXiv  Detail & Related papers  (2025-08-25T20:20:44Z)
• CodeEvo: Interaction-Driven Synthesis of Code-centric Data through Hybrid and Iterative Feedback [21.627909324788597]
  Acquiring high-quality instruction-code pairs is essential for training Large Language Models.<n>We propose CodeEvo, a framework that synthesizes code data through iterative interactions between two LLM agents.
  arXiv  Detail & Related papers  (2025-07-25T16:12:51Z)
• Towards A Generalist Code Embedding Model Based On Massive Data Synthesis [35.04242699869519]
  We introduce textbfCodeR (underlineCode underlineRetrieval), a state-of-the-art embedding model for general-purpose code retrieval.<n>The superior performance of CodeR is built upon CodeR-Pile, a large-scale synthetic dataset constructed under the DRU principle.
  arXiv  Detail & Related papers  (2025-05-19T04:37:53Z)
• KodCode: A Diverse, Challenging, and Verifiable Synthetic Dataset for Coding [49.56049319037421]
  KodCode is a synthetic dataset that addresses the persistent challenge of acquiring high-quality, verifiable training data.<n>It comprises question-solution-test triplets that are systematically validated via a self-verification procedure.<n>This pipeline yields a large-scale, robust and diverse coding dataset.
  arXiv  Detail & Related papers  (2025-03-04T19:17:36Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.