{"ID":5937037,"CreatedAt":"2026-07-07T03:14:33.014478982Z","UpdatedAt":"2026-07-09T14:33:30.924921582Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2607.05150","arxiv_id":"2607.05150","title":"Claim-Level Rubric Rewards for Video Caption Reinforcement Learning","abstract":"In this paper, we introduce Claim-Level Rubric Rewards (CuRe), a structured reward framework designed to address the reward-design bottleneck in reinforcement learning for dense video captioning. Existing reward designs generally fall into two categories: holistic response-level judgment across heterogeneous criteria, or alignment-based evaluation against reference captions. However, both paradigms suffer from fundamental limitations. Holistic rewards struggle to ensure factual accuracy and are prone to stylistic reward hacking, while reference-based rewards overly rely on rigid textual alignment, failing to preserve the completeness and diversity inherent to open-ended generation tasks. To address these challenges, CuRe reformulates reward modeling as fine-grained claim-level verification. Specifically, CuRe decomposes captions into category-aware atomic claims through a structured rubric, converting holistic evaluation into simpler and more reliable claim-level verification.","short_abstract":"In this paper, we introduce Claim-Level Rubric Rewards (CuRe), a structured reward framework designed to address the reward-design bottleneck in reinforcement learning for dense video captioning. Existing reward designs generally fall into two categories: holistic response-level judgment across heterogeneous criteria,...","url_abs":"https://arxiv.org/abs/2607.05150","url_pdf":"https://arxiv.org/pdf/2607.05150v1","authors":"[\"Mingqi Gao\",\"Hongyuan Dong\",\"Yifei Chen\",\"Zhisheng Zhong\",\"Zheng Ruan\",\"Wenjin Hou\",\"Yu Chen\",\"Han Hu\",\"Yansong Tang\"]","published":"2026-07-06T14:33:41Z","proceeding":"cs.CV","tasks":"[\"cs.CV\"]","methods":"[\"Reinforcement Learning\"]","has_code":false}