mlsanityData to 14 Oct 2026

Robust Code RL via Faulty-Code-Driven Test case Synthesis and Dense Reward Shaping

arXiv 2608.24135 ↗First posted 25 Aug 2026cs.AI cs.SEcited by 0

On the map: LLM alignment / RLHF / preference optimisation → Process reward models for reasoning

Abstract (arXiv)

Reinforcement Learning from Verifiable Rewards (RLVR) is pivotal for enhancing LLM code generation, yet its efficacy is often hindered by insufficient test case coverage, leading to reward hacking and policy degradation. To address this, we propose RobustTests, a framework featuring a faulty-code-driven test case synthesis strategy. By leveraging "near-correct" faulty codes, RobustTests captures latent logical discrepancies and employs validator agents with behavioral feature clustering to filter invalid or redundant test cases. Additionally, a stepwise dense reward function based on pass rates is introduced to mitigate false negatives and enhance training robustness. Using this pipeline, we construct an augmented version of the CodeContests+ dataset with superior diagnostic utility. Experimental results show that RL fine-tuning of Qwen3-32B via RobustTests achieves a 3% absolute gain on LiveCodeBench, demonstrating its effectiveness in advancing LLM code generation proficiency. Codes and data are available at https://huggingface.co/datasets/sid6/RobustTests.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Process reward models for reasoning, one of 10 directions in LLM alignment / RLHF / preference optimisation. Builds on 19 of its 34 references that are in the map. First posted on arXiv 25 Aug 2026.

Builds on

19 of 34 references.

Closest ancestors · 2

Same direction, newest first

  1. 2025ACECODER: Acing Coder RL via Automated Test-Case Synthesis3 Feb
  2. 2025Dynamic Scaling of Unit Tests for Code Reward Modeling2 Jan

Other references · 17

  1. 2022Training language models to follow instructions with human feedbackcited by 15,971
  2. 2017Proximal Policy Optimization Algorithmscited by 15,215
  3. 2021Program Synthesis with Large Language Modelscited by 3,637
  4. 2024LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Codecited by 2,004
  5. 2022Competition-Level Code Generation with AlphaCodecited by 1,462
  6. 2023Is Your Code Generated by ChatGPT Really Correct? Rigorous Evaluation of Large Language Models for Code Generationcited by 1,391
  7. 2024A Survey on Large Language Models for Code Generationcited by 658
  8. 2025Rubrics as Rewards: Reinforcement Learning Beyond Verifiable Domainscited by 366
  9. 2025SWE-smith: Scaling Data for Software Engineering Agentscited by 285
  10. 2023Is Self-Repair a Silver Bullet for Code Generation?cited by 197
  11. 2025KodCode: A Diverse, Challenging, and Verifiable Synthetic Dataset for Codingcited by 119
  12. 2025Every Step Evolves: Scaling Reinforcement Learning for Trillion-Scale Thinking Modelcited by 55
5 more
  1. 2025CodeContests+: High-Quality Test Case Generation for Competitive Programmingcited by 46
  2. 2025Learning to Solve and Verify: A Self-Play Framework for Code and Test Generationcited by 35
  3. 2024Large Language Models as Test Case Generators: Performance Evaluation and Enhancementcited by 28
  4. 2024Catastrophic Goodhart: regularizing RLHF with KL divergence does not mitigate heavy-tailed reward misspecificationcited by 23
  5. 2025HardTests: Synthesizing High-Quality Test Cases for LLM Codingcited by 23

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025Rethinking Verification for LLM Code Generation: From Generation to Testing
  2. 2025Klear-CodeTest: Scalable Test Case Generation for Code Reinforcement Learning
  3. 2026CVeDRL: An Efficient Code Verifier via Difficulty-aware Reinforcement Learning
  4. 2026Two-Stage Reinforcement Learning for Sound and Adversarial Test Generation in Code LLMs
  5. 2026CodeScaler: Scaling Code LLM Training and Test-Time Inference via Reward Models
  6. 2026EvolveCoder: Evolving Test Cases via Adversarial Verification for Code Reinforcement Learning
  7. 2026Code-A1: Adversarial Evolving of Code LLM and Test LLM via Reinforcement Learning
  8. 2026Improving Small Language Models for Code Generation with Reinforcement Learning from Verification Feedback