发表机构
Beijing Institute of Technology; China Meteorological Administration; Tsinghua University; Singapore Management University（北京理工大学; 中国气象局; 清华大学; 新加坡管理大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对生成式推荐中波束外有效标识符的失败，提出超越波束方法，通过积分流修复分配并补全候选，显著提升召回与排名指标。
AI 中文摘要
生成式推荐器通过生成标识符来检索物品，但在目录扩展后，一个有效的标识符可能仍然留在波束之外。这引发了两个相互关联的问题：标识符分配修复能解决哪些失败，以及检索应如何在初始波束之外进行？我们使用固定生成器和保留旧标识符来刻画分配修复。输出不变性证书识别出所有可接受分配共有的失败。在共同的生效前缀下，耦合的支持和排名约束给出了目标恢复的新物品数量的精确可行区间。基于这一刻画，超越波束（BB）通过积分流公式获得最小替换修复，选择共享映射并调整生成器。在推理时，生成式似然和协作证据定义了一个用于排名、候选优先级和停止的分数。保留的前缀边界指导候选补全，并在满足停止条件时证明其全局Top-K。对有限目录的全面评估确认了在每个可行情况下的构造。在三个Amazon Reviews类别和三个随机种子中，完整的T5流程在平均Recall@10上提高了15.5--46.3%，在NDCG@10上提高了15.2--44.4%，相对于每个数据集和指标上评估的最佳生成式基线。匹配对照表明，共享构造和调整在Beauty和Toys上改善了新目标的排名和认证效率。组合评分和候选补全在T5和仅解码器LC-Rec上均提高了所有三个数据集的NDCG@10。
英文摘要
Generative recommenders retrieve items by generating identifiers, but a valid identifier can remain outside the beam after catalog expansion. This raises two connected questions: which failures can identifier assignment repair, and how should retrieval proceed beyond the initial beam? We characterize assignment repair with a fixed generator and retained old identifiers. Output-invariance certificates identify failures shared by all admissible assignments. Under a common effective prefix, coupled support and ranking constraints give the exact feasible interval of new-item counts for target recovery. Building on this characterization, Beyond the Beam (BB) obtains minimum-replacement repairs through an integral flow formulation, selects a shared map and adapts the generator. At inference, generative likelihood and collaborative evidence define one score for ranking, candidate priority and stopping. Retained prefix bounds guide candidate completion and certify its global Top-$K$ when the stopping condition is met. Exhaustive finite-catalog evaluation confirms construction in every feasible case. Across three Amazon Reviews categories and three random seeds, the full T5 procedure improves mean Recall@10 by 15.5--46.3% and NDCG@10 by 15.2--44.4% over the best-performing evaluated generative baseline for each dataset and metric. Matched controls show that shared construction and adaptation improve new-target ranking and certification efficiency on Beauty and Toys. Combined scoring and candidate completion improve NDCG@10 across all three datasets with both T5 and decoder-only LC-Rec.
Comments51 pages, 14 figures, including appendices
