🛰️ Daily AI Frontier
‹ back to 2026-09-09

Entropy-Regularized Rank-Masked Policy Optimization for Test-Time Reinforcement Learning in Code Generation

Research LLMs & Foundation Models

Ranking

Overall 74
Content 90
Popularity 37

Observed public metrics from 1 member.

Representative image for Entropy-Regularized Rank-Masked Policy Optimization for Test-Time Reinforcement Learning in Code Generation

Merged summary

TL;DR - ERPO enables test-time reinforcement learning for code generation by rewarding behavioral agreement among candidate programs on automatically constructed probe inputs. It also limits reward hacking and policy drift, improving pass@1 and pass@k across in-domain adaptation and zero-shot transfer benchmarks.

  • Probe-driven TTRL generates output-free test inputs from problem statements and executes candidate programs on them.
  • Probe Consensus Reward compares program behavior rather than source-code surface form, providing a training signal without canonical answers.
  • Rank masking makes conservative negative updates when consensus is low, while an entropy ceiling constrains policy drift.
  • Behavioral consensus remains an imperfect verifier, so ERPO explicitly addresses the risk of spurious agreement.

Sources (1)

Entropy-Regularized Rank-Masked Policy Optimization for Test-Time Reinforcement Learning in Code Generation

arXiv cs.LG Jiacheng Xu, Feng Chen, Xiuneng Xu, Bo An 2026-09-08 arXiv:2609.09135
Public signals Semantic Scholar citations 0 · Semantic Scholar influential citations 0
Providers: Hugging Face · N/A OpenAlex · N/A Publisher · N/A Semantic Scholar · Citations 0 · Influential citations 0 X · N/A Fetched 2026-09-21 14:23:14.529929 UTC

TL;DR - ERPO enables test-time reinforcement learning for code generation by rewarding behavioral agreement among candidate programs on automatically constructed probe inputs. It also limits reward hacking and policy drift, improving pass@1 and pass@k across in-domain adaptation and zero-shot transfer benchmarks.

  • Probe-driven TTRL generates output-free test inputs from problem statements and executes candidate programs on them.
  • Probe Consensus Reward compares program behavior rather than source-code surface form, providing a training signal without canonical answers.
  • Rank masking makes conservative negative updates when consensus is low, while an entropy ceiling constrains policy drift.
  • Behavioral consensus remains an imperfect verifier, so ERPO explicitly addresses the risk of spurious agreement.
item →