用奖励信号优化检索模型，避免破坏语义结构。
Learning from What You Retrieve: Online RL Fine-Tuning for Semantic Retrieval
- 只对高回报检索项更新梯度，保持嵌入空间稳定
- 在工业级数据集上显著优于标准强化学习方法
- 适合需要冻结索引的线上检索系统
在大规模电商检索中，双编码器检索器通常通过对比相似性进行优化，而下游重排序器则捕捉更细粒度的相关性偏好；这种目标不匹配限制了端到端检索质量。强化学习可通过奖励模型反馈来适应检索器，但我们发现标准策略梯度更新会破坏嵌入几何结构，尤其在文档索引因工业约束必须冻结时。为此，我们提出PAO（Positive-Advantage-Only），一种选择性强化学习优化方法。分析表明，在固定高维空间中无差别地惩罚负样本（推远）会破坏预训练的语义流形。PAO仅对具有正优势的检索项应用梯度更新，有效将查询嵌入拉向高奖励区域，同时保持全局拓扑稳定性。在大规模工业数据集和公开基准上的实验表明，PAO显著优于标准强化学习与蒸馏基线。
原文摘要 · Abstract (English)
In large-scale e-commerce retrieval, dual-encoder retrievers are op- timized for contrastive similarity, whereas downstream rerankers capture finer-grained relevance preferences; this objective mis- match limits end-to-end retrieval quality. Reinforcement Learning offers a way to use reward-model feedback for retriever adaptation, but we observe that standard policy-gradient updates can degrade embedding geometry, especially when the document index must remain frozen due to industrial constraints. To address this, we propose PAO (Positive-Advantage-Only), a selective RL optimization method. Our analysis reveals that in- discriminate penalization of negative samples (pushing away) in a frozen high-dimensional space disrupts pre-trained semantic man- ifolds. PAO selectively applies gradient updates only to retrieved items with positive advantages, effectively pulling query embed- dings toward high-reward regions while preserving global topo- logical stability. Experiments on both a massive industrial dataset and public benchmarks demonstrate that PAO significantly outper- forms standard RL and distillation baselines.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。