面试题库

全部 Django Elasticsearch FastAPI Flask Function Call JVM Java IO Java 后端 Java 基础 Java 并发 Java 新特性 Java 集合 JavaWeb Kafka LLM LangChain LangGraph MongoDB MyBatis MySQL Netty Nginx Python RAG RabbitMQ Redis RocketMQ Spring Spring AI SpringBoot SpringCloud SpringMVC ZooKeeper 分布式后端技术大厂面经大数据微服务提示词工程操作系统数据库数据结构智能体海康威视2026秋季社招专项消息队列监控运维算法系统设计缓存计算机基础计算机网络设计模式面试题高可用高并发

RAG 中等

What challenges do re-rankers face regarding computational overhead and latency?

作答 →

RAG 中等

Compare the noise reduction capabilities of re-rankers versus simply increasing the similarity threshold in initial retrieval. When would each approach be more appropriate?

作答 →

RAG 中等

What is a VectorDB, and how is it utilized in RAG retrieval?

作答 →

RAG 中等

What are the key considerations when choosing an embedding model for a RAG system?

作答 →

RAG 中等

What are embeddings, and how are they utilized in RAG retrieval?

作答 →

RAG 中等

What does context precision measure in a RAG retriever, and how does it differ from context recall?

作答 →

RAG 中等

Compare Context Recall with traditional information retrieval recall. Why is Context Recall computed using "ground truth claims" rather than simply counting relevant documents?

作答 →

RAG 中等

Explain the possible reasons for a RAG retrieval system with consistently low context precision

作答 →

RAG 中等

How do you handle ambiguous or vague user queries in RAG systems?

作答 →

RAG 中等

What happens with a weak generator LLM in a RAG system?

作答 →

RAG 中等

Compare reasoning vs. non-reasoning LLMs for RAG systems

作答 →

RAG 中等

What are the key hyperparameters in a RAG pipeline?

作答 →