ScaleBox: Enabling High-Fidelity and Scalable Code Verification for Large Language Models

• URL: http://arxiv.org/abs/2604.27467v1
• Date: Thu, 30 Apr 2026 06:09:17 GMT
• Title: ScaleBox: Enabling High-Fidelity and Scalable Code Verification for Large Language Models
• Abstract summary: Code sandboxes have emerged as a critical infrastructure for advancing the coding capabilities of large language models.<n>Existing systems fail to provide accurate verification and efficiency under high-concurrency workloads.<n>We present ScaleBox, a high-fidelity and scalable system designed to address these limitations in large-scale code training.
• Score: 65.56970356058655
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Code sandboxes have emerged as a critical infrastructure for advancing the coding capabilities of large language models, providing verifiable feedback for both RL training and evaluation. However, existing systems fail to provide accurate verification and efficiency under high-concurrency workloads. We present ScaleBox, a high-fidelity and scalable system designed to address these limitations in large-scale code training. ScaleBox introduces automated special-judge generation and management, fine-grained parallel execution across test cases with seamless multi-node coordination, and a configuration-driven evaluation suite for reproducible benchmarking. A series of experiments demonstrates that ScaleBox significantly enhances code verification accuracy and efficiency. Our further RLVR experiments show that ScaleBox substantially improves both performance on LiveCodeBench and training stability, significantly outperforming heuristic-matching baselines. By providing a reliable and high-throughput infrastructure, ScaleBox facilitates more effective research and development in large-scale code training.

Related papers

• When Do Intrinsic Rewards Work for Code Reasoning? A Comprehensive Study [77.64957118012423]
  Reinforcement learning with verifiable rewards (RLVR) has driven substantial progress in large language model reasoning.<n>Recent work addresses this by deriving rewards from a model's own signals, such as majority voting or confidence-based scores.<n>We present a systematic empirical study of intrinsic reward methods for code generation.
  arXiv  Detail & Related papers  (2026-06-18T19:15:50Z)
• LocateAnything: Fast and High-Quality Vision-Language Grounding with Parallel Box Decoding [80.49137401046201]
  We introduce a unified generative grounding and detection framework based on Parallel Box Decoding (PBD)<n>We show that PBD improves both decoding throughput and localization accuracy.<n>We further develop a scalable data engine and curate LocateAnything-Data, a large-scale dataset with more than 138 million training samples.
  arXiv  Detail & Related papers  (2026-05-26T17:59:12Z)
• CodeScaler: Scaling Code LLM Training and Test-Time Inference via Execution-Free Reward Models [32.910307078704996]
  Execution-free reward model designed to scale both reinforcement learning training and test-time inference for code generation.<n>CodeScaler improves Qwen3-8B-Base by an average of +11.72 points, outperforming binary execution-based RL by +1.82 points.
  arXiv  Detail & Related papers  (2026-02-04T17:56:00Z)
• Aletheia: What Makes RLVR For Code Verifiers Tick? [51.371034079170435]
  Verifiers trained via Reinforcement Learning from Verifiable Rewards (RLVR) are a prominent fixture of the Large Language Model (LLM) post-training pipeline.<n>Code verifiers remain valuable toward judging model outputs in scenarios where execution feedback is hard to obtain.<n>We examine components of the RLVR-based verifier training recipe widely credited for its success.
  arXiv  Detail & Related papers  (2026-01-17T22:30:45Z)
• WeChat-YATT: A Scalable, Simple, Efficient, and Production Ready Training Library [34.5103280294468]
  We introduce WeChat-YATT Yet Another Transformer Trainer (YATT) in WeChat, a simple, scalable, and balanced RLHF training framework.<n>YATT features a parallel controller programming model that enables flexible and efficient orchestration of complex RLHF.<n>We evaluate WeChat-YATT across diverse experimental scenarios, demonstrating its substantial throughput improvements over state-of-the-art RLHF training frameworks.
  arXiv  Detail & Related papers  (2025-08-11T13:31:53Z)
• Klear-CodeTest: Scalable Test Case Generation for Code Reinforcement Learning [43.30900834053253]
  We present Klear-CodeTest, a comprehensive test case synthesis framework featuring rigorous verification to ensure quality and reliability of test cases.<n>The proposed G-V framework generates comprehensive test cases including both regular and corner cases, enhancing test coverage and discriminative power for solution correctness assessment.<n>In addition, we design a multi-layered security sandbox system optimized for online verification platforms, guaranteeing safe and reliable code execution.
  arXiv  Detail & Related papers  (2025-08-07T07:36:01Z)
• Towards Effective Code-Integrated Reasoning [89.47213509714578]
  We investigate code-integrated reasoning, where models generate code when necessary and integrate feedback by executing it through a code interpreter.<n>Tool-augmented reinforcement learning can still suffer from potential instability in the learning dynamics.<n>We develop enhanced training strategies that balance exploration and stability, progressively building tool-use capabilities while improving reasoning performance.
  arXiv  Detail & Related papers  (2025-05-30T11:30:18Z)
• VerifyBench: Benchmarking Reference-based Reward Systems for Large Language Models [53.128374915958624]
  OpenAI o1 and DeepSeek-R1 have achieved remarkable performance in the domain of reasoning.<n>A key component of their training is the incorporation of verifiable rewards within reinforcement learning.<n>Existing reward benchmarks do not evaluate reference-based reward systems.
  arXiv  Detail & Related papers  (2025-05-21T17:54:43Z)

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.