Open Rubric System: Scaling Reinforcement Learning with Pairwise Adaptive Rubric
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866911472729194496 |
|---|---|
| author | Jia, Ruipeng Yang, Yunyi Wu, Yuxin Gai, Yongbo Tao, Siyuan Zhou, Mengyu Lin, Jianhe Jiang, Xiaoxi Jiang, Guanjun |
| author_facet | Jia, Ruipeng Yang, Yunyi Wu, Yuxin Gai, Yongbo Tao, Siyuan Zhou, Mengyu Lin, Jianhe Jiang, Xiaoxi Jiang, Guanjun |
| contents | Scalar reward models compress multi-dimensional human preferences into a single opaque score, creating an information bottleneck that often leads to brittleness and reward hacking in open-ended alignment. We argue that robust alignment for non-verifiable tasks is fundamentally a principle generalization problem: reward should not be a learned function internalized into a judge, but an explicit reasoning process executed under inspectable principles. To operationalize this view, we present the Open Rubric System (OpenRS), a plug-and-play, rubrics-based LLM-as-a-Judge framework built around Pairwise Adaptive Meta-Rubrics (PAMR) and lightweight Pointwise Verifiable Rubrics (PVRs), which provide both hard-constraint guardrails and verifiable reward components when ground-truth or programmatic checks are available. OpenRS uses an explicit meta-rubric -- a constitution-like specification that governs how rubrics are instantiated, weighted, and enforced -- and instantiates adaptive rubrics on the fly by conditioning on the semantic differences between two candidate responses. It then performs criterion-wise pairwise comparisons and aggregates criterion-level preferences externally, avoiding pointwise weighted scalarization while improving discriminability in open-ended settings. To keep principles consistent yet editable across various domains, we introduce a two-level meta-rubric refinement pipeline (automated evolutionary refinement for general principles and a reproducible human-in-the-loop procedure for domain principles), complemented with pointwise verifiable rubrics that act as both guardrails against degenerate behaviors and a source of verifiable reward for objective sub-tasks. Finally, we instantiate OpenRS as reward supervision in pairwise RL training. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_14069 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Open Rubric System: Scaling Reinforcement Learning with Pairwise Adaptive Rubric Jia, Ruipeng Yang, Yunyi Wu, Yuxin Gai, Yongbo Tao, Siyuan Zhou, Mengyu Lin, Jianhe Jiang, Xiaoxi Jiang, Guanjun Computation and Language Scalar reward models compress multi-dimensional human preferences into a single opaque score, creating an information bottleneck that often leads to brittleness and reward hacking in open-ended alignment. We argue that robust alignment for non-verifiable tasks is fundamentally a principle generalization problem: reward should not be a learned function internalized into a judge, but an explicit reasoning process executed under inspectable principles. To operationalize this view, we present the Open Rubric System (OpenRS), a plug-and-play, rubrics-based LLM-as-a-Judge framework built around Pairwise Adaptive Meta-Rubrics (PAMR) and lightweight Pointwise Verifiable Rubrics (PVRs), which provide both hard-constraint guardrails and verifiable reward components when ground-truth or programmatic checks are available. OpenRS uses an explicit meta-rubric -- a constitution-like specification that governs how rubrics are instantiated, weighted, and enforced -- and instantiates adaptive rubrics on the fly by conditioning on the semantic differences between two candidate responses. It then performs criterion-wise pairwise comparisons and aggregates criterion-level preferences externally, avoiding pointwise weighted scalarization while improving discriminability in open-ended settings. To keep principles consistent yet editable across various domains, we introduce a two-level meta-rubric refinement pipeline (automated evolutionary refinement for general principles and a reproducible human-in-the-loop procedure for domain principles), complemented with pointwise verifiable rubrics that act as both guardrails against degenerate behaviors and a source of verifiable reward for objective sub-tasks. Finally, we instantiate OpenRS as reward supervision in pairwise RL training. |
| title | Open Rubric System: Scaling Reinforcement Learning with Pairwise Adaptive Rubric |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2602.14069 |