RAILS：检索增强的 LLM 大规模增量聚类方法
RAILS: Retrieval-Augmented Incremental LLM Clustering at Scale
RAILS 用 LLM 做文本聚类，六项基准平均成绩超过最强旧方法，还真的在工单系统里替换掉了 HDBSCAN。
LLM 做聚类落到生产环境有两个瓶颈：提示词装不下整个标签空间，逐条串行处理也达不到真实负载的吞吐量。论文提出的 RAILS 把聚类改造成对不断增长的标签池的检索循环，再靠文档批处理和有界并发来扩展规模。在六个公开基准上，RAILS 平均超过此前最强的 LLM 聚类方法：准确率从 51.2% 提到 59.3%，NMI 从 67.2% 提到 74.8%，ARI 从 45.4% 提到 54.7%。团队还在一个 SaaS 工单主题发现流水线里用 RAILS 替换掉了 HDBSCAN 环节，聚类质量更高，支持提示词层面的透明控制，且能带状态做增量运行。
RAILS: Retrieval-Augmented Incremental LLM Clustering at Scale
Using a Large Language Model (LLM) as the clusterer at production scale is hard: prompts cannot hold the entire label space, and per-document serial processing does not deliver the throughput real workloads require. We present RAILS, a retrieval-augmented incremental LLM clusterer that turns clustering into a simple loop over a growing label pool and scales through document batching with bounded concurrency. On six public benchmarks RAILS exceeds the strongest prior LLM-clustering method on average, lifting accuracy from 51.2% to 59.3%, NMI from 67.2% to 74.8%, and ARI from 45.4% to 54.7%. We further report production-deployment evidence from a SaaS ticket-topic-discovery pipeline, where RAILS has replaced a traditional HDBSCAN stage with higher clustering quality, transparent prompt-driven control, and stateful incremental operation.