离线生成解释池，用排序模型快速选最优解，显著降低推荐系统延迟和成本。
Pairwise Ranking Outperforms Single-Action RL for Offline Explanation Selection: A Practical Lesson
- 预先生成解释候选池，请求时仅用轻量级排序器选择，无需实时调用大模型。
- 成对排序方法在2958对数据上达F1=0.500，优于单动作强化学习方法。
- 适用于工业级可解释推荐系统，尤其适合对响应速度和成本敏感的场景。
基于大语言模型的工业级可解释推荐系统面临高昂服务成本：每次请求需触发大模型生成，延迟达数百毫秒且成本随流量线性增长。本文将生成与选择分离：提前生成固定解释候选池（六种提示风格，两种通用大模型），请求时由小型驻内存选择器从池中挑选。该架构无需GPU，响应时间低于100毫秒。主要基准为包含2958对样本的Google Local XRec子集，评估六种离线池选择器（LambdaRank、PPO、GRPO、DPO、师生蒸馏）和三种知识图谱路径选择器（随机游走、边不相交枚举、MMR重排序路径）。以300对MovieLens-1M数据集及Claude-Sonnet-4.5作为内部交叉验证，因该场景无公开基准。所有方法均采用BERTScore-F1协议，五次随机种子平均。LambdaRank在Google Local上达F1=0.500，超过G-Refer和XRec，MovieLens-1M上达F1=0.329。种子方差低于0.003，结果稳定：成对学习排序优于单动作强化学习（如PPO、GRPO、DPO），后者每轮仅使用一个标注候选，浪费其余K-1个标签。知识图谱路径类方法目标不同：三者在Google Local上达成USR=1.000，MovieLens-1M上为0.997–1.000，因每次请求路径接地生成唯一输出，避免缓存大模型输出的模板坍缩问题。对比Claude 3 Haiku与Claude Haiku 4.5的生成器研究显示，F1仅微小变化（0.001–0.006），但选择器排名不变，说明选择器与生成器可独立评估，尽管绝对F1依赖生成器。端到端部署成本接近15美元，运行于通用硬件。
原文摘要 · Abstract (English)
Industrial explainable-recommendation systems built on LLMs incur a substantial serving cost: each request triggers an LLM generation, with latency in the hundreds of milliseconds and cost that scales linearly with traffic. We separate generation from selection: explanations are produced ahead of time as a frozen candidate pool (six prompt styles, two commodity LLMs), and a small CPU-resident selector picks one at request time. The stack needs no GPU and returns in under 100 ms. Our primary benchmark is a 2,958-pair XRec Google Local subset, evaluating six offline-pool selectors (LambdaRank, PPO, GRPO, DPO, teacher-student distillation) and three KG-path selectors (random walks, edge-disjoint enumeration, MMR-reranked paths). A 300-pair MovieLens-1M split with Claude-Sonnet-4.5 references serves as an internal cross-dataset check, since no public benchmark exists for this setting. All variants use the same BERTScore-F1 protocol as XRec and G-Refer, averaged across five seeds. LambdaRank reaches F1 = 0.500 on Google Local, exceeding both G-Refer and XRec, and F1 = 0.329 on the MovieLens-1M check. With seed variance below 0.003 F1, the ordering is reliable: pairwise learning-to-rank outperforms single-action RL (PPO, GRPO, DPO), which use only one labelled candidate per rollout, leaving K-1 labels unused. The KG-path family targets a different objective: all three variants reach USR = 1.000 on Google Local and 0.997-1.000 on MovieLens-1M, since per-request path grounding yields a unique output per query, avoiding template-collapse failures affecting cached-LLM outputs. A generator-pool study comparing Claude 3 Haiku and Claude Haiku 4.5 shows small F1 shifts (0.001-0.006) while preserving selector ranking: selector and generator can be evaluated independently, though absolute F1 depends on the generator. End-to-end build cost is near $15 on commodity hardware.
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。