dorsal/arxiv
View SchemaRubricHub: A Comprehensive and Highly Discriminative Rubric Dataset via Automated Coarse-to-Fine Generation
| Authors | Sunzhu Li, Jiale Zhao, Miteto Wei, Huimin Ren, Yang Zhou, Jingwen Yang, Shunyu Liu, Kaike Zhang, Wei Chen |
|---|---|
| Categories | |
| ArXiv ID | 2601.08430vv1 |
| URL | https://arxiv.org/abs/2601.08430 |
| License | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ |
Abstract
Reinforcement Learning with Verifiable Rewards (RLVR) has driven substantial progress in reasoning-intensive domains like mathematics. However, optimizing open-ended generation remains challenging due to the lack of ground truth. While rubric-based evaluation offers a structured proxy for verification, existing methods suffer from scalability bottlenecks and coarse criteria, resulting in a supervision ceiling effect. To address this, we propose an automated Coarse-to-Fine Rubric Generation framework. By synergizing principle-guided synthesis, multi-model aggregation, and difficulty evolution, our approach produces comprehensive and highly discriminative criteria capable of capturing the subtle nuances. Based on this framework, we introduce RubricHub, a large-scale ($\sim$110k) and multi-domain dataset. We validate its utility through a two-stage post-training pipeline comprising Rubric-based Rejection Sampling Fine-Tuning (RuFT) and Reinforcement Learning (RuRL). Experimental results demonstrate that RubricHub unlocks significant performance gains: our post-trained Qwen3-14B achieves state-of-the-art (SOTA) results on HealthBench (69.3), surpassing proprietary frontier models such as GPT-5. The code and data will be released soon.
{
"annotation_id": "fb492f7b-5fe5-4b4e-bf25-9eb22c2f573b",
"date_created": "2026-02-17T05:53:16.195000Z",
"date_modified": "2026-02-17T05:53:16.195000Z",
"file_hash": "575751968b9d72ecfd67073ebc87e5ea24944c3e74ea1f124f155860833b4871",
"private": false,
"record": {
"abstract": "Reinforcement Learning with Verifiable Rewards (RLVR) has driven substantial progress in reasoning-intensive domains like mathematics. However, optimizing open-ended generation remains challenging due to the lack of ground truth. While rubric-based evaluation offers a structured proxy for verification, existing methods suffer from scalability bottlenecks and coarse criteria, resulting in a supervision ceiling effect. To address this, we propose an automated Coarse-to-Fine Rubric Generation framework. By synergizing principle-guided synthesis, multi-model aggregation, and difficulty evolution, our approach produces comprehensive and highly discriminative criteria capable of capturing the subtle nuances. Based on this framework, we introduce RubricHub, a large-scale ($\\sim$110k) and multi-domain dataset. We validate its utility through a two-stage post-training pipeline comprising Rubric-based Rejection Sampling Fine-Tuning (RuFT) and Reinforcement Learning (RuRL). Experimental results demonstrate that RubricHub unlocks significant performance gains: our post-trained Qwen3-14B achieves state-of-the-art (SOTA) results on HealthBench (69.3), surpassing proprietary frontier models such as GPT-5. The code and data will be released soon.",
"arxiv_id": "2601.08430",
"authors": [
"Sunzhu Li",
"Jiale Zhao",
"Miteto Wei",
"Huimin Ren",
"Yang Zhou",
"Jingwen Yang",
"Shunyu Liu",
"Kaike Zhang",
"Wei Chen"
],
"categories": [
"cs.AI"
],
"license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
"title": "RubricHub: A Comprehensive and Highly Discriminative Rubric Dataset via Automated Coarse-to-Fine Generation",
"url": "https://arxiv.org/abs/2601.08430",
"version": "v1"
},
"schema_id": "dorsal/arxiv",
"source": {
"execution_id": "83e2f26f-61e4-468a-a0d7-8c86f0034d42",
"id": "arXiv Dataset",
"type": "Model",
"variant": "snapshot-2026-01-17",
"version": "0.1.0"
},
"user_id": 1000002
}