ExecuCritic: Calibrated Critic Shaping for Code Generation with Verifiable Rewards

• URL: http://arxiv.org/abs/2609.16604v1
• Date: Tue, 15 Sep 2026 03:58:21 GMT
• Title: ExecuCritic: Calibrated Critic Shaping for Code Generation with Verifiable Rewards
• Abstract summary: Execution feedback is a useful supervision signal for code models because unit tests are objective and directly measure program correctness.<n>We propose ExecuCritic, a joint training framework in which a coder and a critic are updated on the same execution rollouts.
• Score: 2.724374986592201
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Execution feedback is a useful supervision signal for code models because unit tests are objective and directly measure program correctness. Its weakness is that an entire program is often reduced to one pass or fail bit, leaving RLVR to solve a difficult credit assignment problem. At the same time, coding systems often include separate reviewer or tester roles, but these critics are usually prompted rather than trained and are not calibrated against execution. We propose ExecuCritic, a joint training framework in which a coder and a critic are updated on the same execution rollouts. The critic predicts pass or fail outcomes and gives short diagnostic feedback; the coder uses this signal only when the critic agrees with the executor on the current rollout group. Across eight code benchmarks and two recent open backbones, ExecuCritic improves over GRPO without a critic, prompted reviewer systems and scalar reward model baselines, while requiring fewer policy gradient steps and fewer sandbox executions. Ablations and reliability analyses suggest that the gains come from better credit assignment rather than larger sampling budgets.

Related papers

• Benchmarking Code Improvement with Progressive, Adaptive, and Interactive Feedback [13.976149104483449]
  PAIR-Bench is a progressive benchmark for evaluating large language models (LLMs)<n>It transforms an incorrect or incomplete program into a more correct one through feedback-guided refinement.<n>It measures whether a model repairs targeted failures, generalizes beyond the hint, preserves already-correct behavior, and how much assistance it requires.
  arXiv  Detail & Related papers  (2026-07-01T18:20:27Z)
• 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)
• Bias in the Loop: Auditing LLM-as-a-Judge for Software Engineering [0.0]
  Large Language Models are increasingly used as judges to evaluate code artifacts.<n>Current practice lacks a principled account of reliability and bias.<n>This paper studies LLM-as-a-Judge for code through a measurement-first lens.
  arXiv  Detail & Related papers  (2026-04-18T02:35:05Z)
• Vibe-Coding: Feedback-Based Automated Verification with no Human Code Inspection, a Feasibility Study [0.0]
  This paper studies feedback-based automated verification of LLM-generated adaptation managers in Collective Adaptive Systems (CAS)<n>We focus on the key challenges of verification in the loop: how to detect failures of generated code at runtime and how to report them.<n>Our findings suggest that feedback precision is the dominant factor for reliable vibe coding in systems designed by domain experts with no programming skills.
  arXiv  Detail & Related papers  (2026-04-16T10:58:03Z)
• CR-Bench: Evaluating the Real-World Utility of AI Code Review Agents [0.7496422063843831]
  We introduce CR-Bench, a benchmarking dataset, and CR-Evaluator, a fine-grained evaluation pipeline for code review agents.<n>We find that code review agents can exhibit a low signal-to-noise ratio when designed to identify all hidden issues.<n>Our analysis identifies the hidden trade-off between issue resolution and spurious findings, revealing a frontier that constrains effective agent design.
  arXiv  Detail & Related papers  (2026-03-10T21:29:42Z)
• CARE What Fails: Contrastive Anchored-REflection for Verifiable Multimodal [84.71254539482369]
  Group-relative reinforcement learning with verifiable rewards (RLVR) often wastes the most informative data it already has the failures.<n>We present CARE, a failure-centric post-training framework for multimodal reasoning that turns errors into supervision.<n> CARE improves accuracy and training smoothness while explicitly increasing the share of learning signal that comes from failures.
  arXiv  Detail & Related papers  (2025-12-22T16:34:21Z)
• RLAC: Reinforcement Learning with Adversarial Critic for Free-Form Generation Tasks [75.52891348667491]
  Open-ended generation tasks require outputs to satisfy diverse and often implicit task-specific evaluation rubrics.<n>The sheer number of relevant rubrics leads to prohibitively high verification costs and incomplete assessments of a response.<n>We propose Reinforcement Learning with Adrial Critic (RLAC), a post-training approach that addresses these challenges via dynamic rubric verification.
  arXiv  Detail & Related papers  (2025-11-03T17:15:05Z)
• Training Language Model to Critique for Better Refinement [58.73039433159486]
  We introduce textbfRefinement-oriented textbfCritique textbfOptimization (RCO), a novel framework designed to train critic models using refinement signals.<n>RCO uses a feedback loop where critiques, generated by the critic model, guide the actor model in refining its responses.<n>By focusing on critiques that lead to better refinements, RCO eliminates the need for direct critique preference assessment.
  arXiv  Detail & Related papers  (2025-06-27T12:10:57Z)
• CodeCriticBench: A Holistic Code Critique Benchmark for Large Language Models [97.18215355266143]
  We introduce a holistic code critique benchmark for Large Language Models (LLMs) called CodeCriticBench.<n>Specifically, our CodeCriticBench includes two mainstream code tasks (i.e., code generation and code QA) with different difficulties.<n>Besides, the evaluation protocols include basic critique evaluation and advanced critique evaluation for different characteristics.
  arXiv  Detail & Related papers  (2025-02-23T15:36:43Z)
• DOCE: Finding the Sweet Spot for Execution-Based Code Generation [69.5305729627198]
  We propose a comprehensive framework that includes candidate generation, $n$-best reranking, minimum Bayes risk (MBR) decoding, and self-ging as the core components. Our findings highlight the importance of execution-based methods and the difference gap between execution-based and execution-free methods.
  arXiv  Detail & Related papers  (2024-08-25T07:10:36Z)

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.