| title | Rubric-to-Code Credit Assignment for Reinforcement Learning |
|---|---|
| authors | Rui Jin, Jikai Chen, Yihan Chen, Hao Zhou, Demin Zhu, Kaichen Yang, Dong Wang, Chenyi Zhuang |
| date | 2026-08-28 |
| venue | arXiv (preprint, not peer reviewed) |
| id | arXiv:2608.27906 |
| affiliation | Ant Group |
| upvotes | 8 (Hugging Face Daily Papers, 2026-11-01) |
| links | abs pdf hf |
| abstract | Interactive web application generation requires models to produce usable HTML, CSS, and JavaScript applications from natural language requests. Unlike conventional code generation, application quality depends on multiple user-facing functional requirements, each often tied to localized code regions such as event handlers, state updates, DOM fragments, or CSS selectors. Standard GRPO collapses these structured outcomes into a single sequence-level reward and applies the resulting advantage uniformly to all tokens, weakening credit assignment. We propose Rubric-to-Code Credit Assignment (RCCA), a reinforcement learning framework that converts rubric-level functional feedback into localized optimization signals over generated code. RCCA builds training tasks around explicit functional rubrics, uses a hierarchical reward to separate format, source-code, runtime, and functional failures, and aligns evaluator-generated textual attributions with responsible code spans and generated tokens. The resulting model, Ling-RCCA-Flash, scores 41.25 on MiniAppBench, improving Ling-3.0-Flash by 32.20 points and slightly surpassing Claude Opus 4.5. It also reaches 76.19 on ArtifactsBench, improving the SFT model by 4.48 points and establishing a new top score under the official ArtifactsBench leaderboard setting by surpassing the GPT-5 score by 3.64 points, suggesting transferable implementation-level gains. |
cite
@misc{jin2026rubric,
title={Rubric-to-Code Credit Assignment for Reinforcement Learning},
author={Rui Jin and Jikai Chen and Yihan Chen and Hao Zhou and Demin Zhu and Kaichen Yang and Dong Wang and Chenyi Zhuang},
year={2026},
eprint={2608.27906},
archivePrefix={arXiv},
url={https://arxiv.org/abs/2608.27906}
}