14 天 RAG:从检索到可信回答
课程题库 · 57 / 57 道题
还有 92 个标签收起标签
D1 为什么要检索:幻觉、知识截止与长上下文的代价,以及一个纯关键词的最小 RAG
什么时候该用检索增强生成,什么时候该微调,什么时候直接把文档塞进上下文就够了?When should you use retrieval-augmented generation, when should you fine-tune, and when is stuffing the documents into the context window good enough?
国内高频海外高频基础#rag-basics#fine-tuning#long-context🔒 解锁 D1 后查看答题要点。
BM25 里的词频饱和与文档长度归一化分别在解决什么问题?把 k1 和 b 都设成 0 会发生什么?In BM25, what problems do term-frequency saturation and document length normalization each solve? What happens if you set both k1 and b to zero?
国内高频海外高频进阶#bm25#ranking#information-retrieval🔒 解锁 D1 后查看答题要点。
一个检索增强生成系统答错了,你怎么定位是检索的锅还是生成的锅?A retrieval-augmented generation system gave a wrong answer. How do you determine whether retrieval or generation is at fault?
国内高频海外高频进阶#debugging#failure-modes#evaluation🔒 解锁 D1 后查看答题要点。
上下文窗口已经做到上百万 token 了,检索这一步会被淘汰吗?Context windows are now in the millions of tokens. Does that make the retrieval step obsolete?
国内高频海外高频深入#long-context#cost#system-design🔒 解锁 D1 后查看答题要点。
D2 embedding 与向量检索:相似度、维度与模型选型,把文本存进 pgvector
余弦相似度和内积什么时候等价?如果向量没有归一化,用内积排序会出什么问题?When are cosine similarity and inner product equivalent? What goes wrong if you rank by inner product on vectors that are not normalized?
国内高频海外高频基础#embeddings#similarity#normalisation🔒 解锁 D2 后查看答题要点。
把 embedding 维度从 1536 降到 512,你会损失什么?什么场景下这个损失可以接受?What do you lose when you cut embedding dimensions from 1536 to 512, and when is that loss acceptable?
国内高频海外高频进阶#embeddings#dimensions#cost🔒 解锁 D2 后查看答题要点。
为什么有些 embedding 模型要求查询和文档加不同的前缀?不加会怎样,你怎么在上线前发现这个问题?Why do some embedding models require different prefixes for queries and documents? What happens if you skip them, and how would you catch it before shipping?
国内高频海外高频进阶#embeddings#model-selection#evaluation🔒 解锁 D2 后查看答题要点。
向量检索能完全取代关键词检索吗?举一个向量必然失手的查询,并说说你会怎么补。Can vector search fully replace keyword search? Give a query where vectors are bound to fail, and say how you would fix it.
国内高频海外高频进阶#hybrid-search#embeddings#retrieval-failure🔒 解锁 D2 后查看答题要点。
D3 文档进来这一关:PDF 与 HTML 解析、表格与扫描件、清洗规则和必须留下的元数据
一份 PDF 解析出来的文字顺序是乱的,你会怎么排查和修复?The text extracted from a PDF comes out in the wrong order. How do you diagnose and fix it?
国内高频海外高频进阶#pdf-parsing#ingestion#data-quality🔒 解锁 D3 后查看答题要点。
文档解析阶段应该保留哪些元数据?少了其中某一项会在哪个环节出问题?Which metadata should a document parsing stage preserve, and which downstream feature breaks if you drop each one?
国内高频海外高频进阶#metadata#ingestion#access-control🔒 解锁 D3 后查看答题要点。
扫描件走光学字符识别之后错字率不低,这些噪声会怎样影响检索和生成?怎么缓解?OCR output from scanned documents carries a non-trivial error rate. How does that noise propagate into retrieval and generation, and how do you mitigate it?
国内高频海外高频深入#ocr#data-quality#hybrid-search🔒 解锁 D3 后查看答题要点。
为什么说解析质量决定了检索质量的上限?举一个具体的传导链条。Why is parsing quality the ceiling on retrieval quality? Walk through one concrete chain of propagation.
国内高频海外高频基础#ingestion#data-quality#failure-analysis🔒 解锁 D3 后查看答题要点。
D4 切块策略:固定、递归、按结构、父子与语义五种切法,以及用评估而不是直觉来选
你怎么决定切块大小?说出你会看的两个指标和一个反例。How do you decide on chunk size? Name two metrics you would look at, and one counterexample.
国内高频海外高频进阶#chunking#evaluation🔒 解锁 D4 后查看答题要点。
父子切块的收益是什么?它在什么情况下反而会拖慢系统?What does parent-child chunking buy you, and when does it slow the system down instead?
国内高频海外高频进阶#chunking#parent-child🔒 解锁 D4 后查看答题要点。
重叠区设成块长的百分之多少合适?重叠过大会带来什么具体问题?What overlap ratio would you use, and what concretely goes wrong when the overlap is too large?
国内高频海外高频基础#chunking#overlap🔒 解锁 D4 后查看答题要点。
语义切分比递归切分贵不少,你怎么向团队证明这笔钱值得花?Semantic chunking costs considerably more than recursive splitting. How would you prove to your team that the money is well spent?
国内高频海外高频深入#chunking#evaluation#cost🔒 解锁 D4 后查看答题要点。
D5 向量索引与库选型:HNSW 与倒排文件、量化省内存、带过滤的查询与多租户隔离
分层可导航小世界图和倒排文件索引你会怎么选?各说一个必须选它的场景,以及各自最该调的参数。How do you choose between an HNSW index and an IVFFlat index? Give one scenario that forces each choice, and name the parameter you would tune first in each.
国内高频海外高频进阶#vector-index#hnsw#ivfflat🔒 解锁 D5 后查看答题要点。
为什么加了 WHERE 条件的向量检索会漏结果?有哪几种修法,代价分别是什么?Why does a vector search with a WHERE clause return fewer results than expected, and what are the fixes and their costs?
国内高频海外高频深入#filtering#iterative-scan#recall🔒 解锁 D5 后查看答题要点。
把向量从全精度换成半精度或二值量化,你会用什么方法确认召回没有明显下降?If you switch your vectors from full precision to half precision or binary quantisation, how do you verify that recall has not dropped materially?
国内高频海外高频进阶#quantization#evaluation#recall🔒 解锁 D5 后查看答题要点。
什么时候应该把向量搬出 PostgreSQL?给出可量化的触发条件,也说说不该搬的理由。When should you move your vectors out of PostgreSQL into a dedicated vector database? Give measurable triggers, and also make the case for staying.
国内高频海外高频进阶#vector-database#architecture#trade-offs🔒 解锁 D5 后查看答题要点。
D6 生成这一侧:上下文怎么排、引用怎么标、什么时候必须拒答,以及流式回答
怎么让模型的引用是真的而不是编的?说出一个不依赖模型自觉的方案。How do you make sure a model's citations are real rather than fabricated? Describe a scheme that does not rely on the model behaving well.
国内高频海外高频进阶#citation-verification#grounding#hallucination🔒 解锁 D6 后查看答题要点。
上下文里材料的排列顺序会影响回答质量吗?如果会,你会怎么排?Does the ordering of retrieved passages in the context affect answer quality? If so, how would you order them?
国内高频海外高频基础#context-assembly#prompt-engineering#ordering🔒 解锁 D6 后查看答题要点。
知识库问答的拒答阈值怎么定?定高了和定低了各自的代价是什么?How do you set the refusal threshold for a knowledge-base assistant, and what does it cost you when the threshold is too high or too low?
国内高频海外高频深入#refusal#thresholds#evaluation🔒 解锁 D6 后查看答题要点。
流式输出的场景下,你怎么保证吐出去的内容不会因为引用校验失败而需要撤回?In a streaming setup, how do you make sure nothing you have already sent needs to be retracted because its citation failed verification?
国内高频海外高频深入#streaming#citation-verification#api-design🔒 解锁 D6 后查看答题要点。
D7 第一周综合:把六天的零件装成一个可一键启动的检索问答服务并复盘
你会怎么划分一个检索增强生成系统的模块边界?其中哪一层最应该做成可替换的,为什么?How would you draw the module boundaries of a RAG system, and which layer most needs to be swappable? Why?
国内高频海外高频基础#architecture#modularity#embeddings🔒 解锁 D7 后查看答题要点。
摄取链路和查询链路应该共享哪些代码?强行复用会带来什么具体问题?What should the ingestion path and the query path share, and what concretely goes wrong when you over-share?
国内高频海外高频进阶#architecture#ingestion#retrieval🔒 解锁 D7 后查看答题要点。
一个检索问答服务上线前你会做哪三项检查?为什么偏偏是这三项?What three checks would you run before shipping a retrieval QA service, and why those three?
国内高频海外高频进阶#production-readiness#citations#refusal🔒 解锁 D7 后查看答题要点。
你刚拼出来的这个检索问答系统,现在最大的风险在哪里?你打算怎么证明这个判断?What is the biggest risk in the RAG service you just assembled, and how would you prove that judgment?
国内高频海外高频深入#evaluation#risk-assessment#retrospective🔒 解锁 D7 后查看答题要点。
D8 评估先行:搭 golden set、算召回与排序指标、用模型当裁判判忠实度
让你从零给一个公司知识库的 RAG 系统建评估集,你会怎么做?多少题才算够用?You need to build an evaluation set from scratch for a RAG system over a company knowledge base. How would you do it, and how many questions are enough?
国内高频海外高频进阶#evaluation#golden-set#rag🔒 解锁 D8 后查看答题要点。
召回率、平均倒数排名、归一化折损累计增益,这三个检索指标分别在什么故障下会先掉下来?只盯一个会漏掉什么?Recall, mean reciprocal rank, and normalized discounted cumulative gain - which failure mode does each one catch first, and what do you miss by watching only one?
国内高频海外高频进阶#retrieval-metrics#evaluation#ranking🔒 解锁 D8 后查看答题要点。
用模型当裁判来评 RAG 的忠实度,有哪些系统性偏差?你怎么发现它们、又怎么证明你的裁判可信?What systematic biases does an LLM judge have when scoring RAG faithfulness, and how do you detect them and prove your judge is trustworthy?
国内高频海外高频深入#llm-as-judge#evaluation#faithfulness🔒 解锁 D8 后查看答题要点。
RAG 的评估集里为什么一定要放语料里没有答案的问题?不放会掩盖什么?Why must a RAG evaluation set include questions the corpus cannot answer, and what does leaving them out hide?
国内高频海外高频基础#evaluation#abstention#golden-set🔒 解锁 D8 后查看答题要点。
D9 混合检索与重排:两路召回、倒数排名融合,再用交叉编码器把前几名重新排一遍
混合检索为什么普遍用倒数排名融合,而不是把两路分数归一化之后加权相加?加权那条路在什么情况下会失控?Why do hybrid retrieval systems usually use reciprocal rank fusion instead of normalizing both scores and adding them with weights? When does the weighted approach break down?
国内高频海外高频进阶#hybrid-search#rank-fusion🔒 解锁 D9 后查看答题要点。
交叉编码器为什么比双编码器准?既然更准,为什么不干脆拿它直接检索全库?Why is a cross-encoder more accurate than a bi-encoder? And if it is more accurate, why not just use it to search the whole corpus directly?
国内高频海外高频基础#cross-encoder#bi-encoder🔒 解锁 D9 后查看答题要点。
你把纯向量检索换成了混合检索加重排,上线之后评估指标反而掉了。你会怎么排查?You replaced pure vector retrieval with hybrid search plus reranking, and after shipping it your eval metrics went down. How do you investigate?
国内高频海外高频深入#hybrid-search#evaluation🔒 解锁 D9 后查看答题要点。
加上重排之后每次提问多了两百毫秒延迟,还多了一笔按次计费的开销。你怎么判断这笔钱该不该付?Adding a reranker costs you 200 ms of extra latency per question plus a per-search fee. How do you decide whether that spend is worth it?
国内高频海外高频深入#rerank#cost-tradeoff🔒 解锁 D9 后查看答题要点。
D10 查询侧优化:改写、假设文档嵌入、多路查询、后退提问与意图路由
假设文档嵌入(HyDE)为什么有效?它在什么情况下会把检索带偏?Why does HyDE (hypothetical document embeddings) work, and when does it steer retrieval in the wrong direction?
国内高频海外高频进阶#hyde#query-transformation#retrieval-quality🔒 解锁 D10 后查看答题要点。
多轮对话里怎么处理指代?不做指代消解最典型的翻车场景是什么?How do you handle coreference in multi-turn RAG, and what is the classic failure when you skip it?
国内高频海外高频基础#coreference#multi-turn#query-rewriting🔒 解锁 D10 后查看答题要点。
上线查询改写之后每问多了一次模型调用,端到端延迟涨了一倍,你怎么判断这笔开销值不值?Query rewriting adds a model call per question and doubles end-to-end latency. How do you decide whether it is worth paying?
国内高频海外高频深入#cost-tradeoff#latency#query-rewriting🔒 解锁 D10 后查看答题要点。
意图路由判错了会怎样?你会怎么设计兜底?What happens when intent routing misclassifies, and how would you design the fallback?
国内高频海外高频进阶#intent-routing#fallback#observability🔒 解锁 D10 后查看答题要点。
D11 高级索引:父子文档、摘要索引、上下文检索,以及树状聚合与图检索的取舍
父子索引和上下文检索都在补『块被切碎』这个问题,它们的差别到底在哪?Parent-child indexing and contextual retrieval both patch the same problem — chunks losing their context. What actually distinguishes them?
国内高频海外高频进阶#indexing#contextual-retrieval#chunking🔒 解锁 D11 后查看答题要点。
上下文检索要给每个块调一次模型,这笔一次性成本怎么估?有哪些办法能压下来?Contextual retrieval needs one model call per chunk. How do you estimate that one-off cost, and what levers bring it down?
国内高频海外高频深入#contextual-retrieval#prompt-caching#cost🔒 解锁 D11 后查看答题要点。
什么样的问题必须上图检索?给一个该上的具体例子和一个不该上的例子。What kind of question actually requires graph retrieval? Give one concrete case where it is justified and one where it is not.
国内高频海外高频深入#graph-rag#multi-hop#cost🔒 解锁 D11 后查看答题要点。
同一份语料建了三套索引,检索时你怎么决定走哪一套?You have built three different indexes over the same corpus. How do you decide which one a query goes to?
国内高频海外高频进阶#index-routing#evaluation#architecture🔒 解锁 D11 后查看答题要点。
D12 Agentic RAG:把检索做成工具,让模型自己决定查不查、查几次、要不要推翻重来
把检索包成一个工具交给模型,这个工具的描述该怎么写?写不好会导致哪些具体的错误行为?You are exposing retrieval to a model as a tool. How do you write the tool description, and what concrete failure modes appear when you write it badly?
国内高频海外高频基础#tool-design#agentic-rag#prompting🔒 解锁 D12 后查看答题要点。
自反思式检索会反复改写查询重试。你怎么保证它一定会停下来,而不是在同一个查询上原地打转?Self-reflective retrieval rewrites the query and retries. How do you guarantee it terminates instead of spinning on the same query forever?
国内高频海外高频进阶#agentic-rag#self-reflection#reliability🔒 解锁 D12 后查看答题要点。
多跳检索里第一跳查错了,后面全跟着错。你会怎么设计容错?In multi-hop retrieval a wrong first hop poisons every hop after it. How would you design for that?
国内高频海外高频深入#multi-hop#error-propagation#agentic-rag🔒 解锁 D12 后查看答题要点。
什么情况下你会拒绝把一个 RAG 系统做成 Agentic 的?拿什么数据说服你的团队?When would you refuse to make a RAG system agentic, and what data would you use to convince your team?
国内高频海外高频进阶#agentic-rag#cost#engineering-judgement🔒 解锁 D12 后查看答题要点。
D13 上生产:增量同步与去重、按权限过滤、缓存分层、链路追踪与成本延迟账
文档更新之后,你怎么做到只重算受影响的块?被删掉的文档又怎么保证一定从索引里消失?After a document changes, how do you recompute only the affected chunks? And how do you guarantee a deleted document really disappears from the index?
国内高频海外高频进阶#incremental-sync#content-hash#index-maintenance🔒 解锁 D13 后查看答题要点。
为什么权限过滤不能放在生成阶段做?放在那里会泄露什么?Why can't access control be applied at the generation step? What exactly leaks if you put it there?
国内高频海外高频深入#access-control#filter-pushdown#multi-tenancy🔒 解锁 D13 后查看答题要点。
RAG 系统里有哪些东西可以缓存?各自的失效条件是什么?What can be cached in a RAG system, and what are the invalidation conditions for each?
国内高频海外高频进阶#caching#invalidation#cost-optimization🔒 解锁 D13 后查看答题要点。
要换一个 embedding 模型,线上系统怎么迁移才能不停机也不掉召回?You need to switch embedding models. How do you migrate a live system without downtime and without losing recall?
国内高频海外高频深入#embedding-migration#zero-downtime#rollout🔒 解锁 D13 后查看答题要点。
D14 综合项目与复盘:多租户企业知识库问答,一张 RAG 决策地图与面试专题
给你一个五百万文档、要求秒级响应、准确率优先的知识库场景,你会怎么设计这套系统?You are handed a knowledge base of five million documents that must answer in about a second, with accuracy as the top priority. How would you design it?
国内高频海外高频深入#system-design#scaling#latency-budget🔒 解锁 D14 后查看答题要点。
你做过的这个 RAG 项目里,哪个决定你现在会改?为什么?Looking back at the RAG project you built, which decision would you change now, and why?
国内高频海外高频深入#retrospective#evidence#chunking🔒 解锁 D14 后查看答题要点。
怎么向不懂技术的业务方证明你的检索系统真的变好了?How do you convince a non-technical stakeholder that your retrieval system actually got better?
国内高频海外高频基础#evaluation#stakeholder-communication#abstention🔒 解锁 D14 后查看答题要点。
RAG 系统上线后用户反馈「答得不准」,你的排查顺序是什么?Users report that your live RAG system 'answers inaccurately'. What is your triage order?
国内高频海外高频进阶#debugging#failure-modes#observability🔒 解锁 D14 后查看答题要点。
如果预算只够做三件事来提升一个已有 RAG 系统的效果,你选哪三件?为什么是这三件?If you could only fund three changes to improve an existing RAG system, which three would you pick and why those three?
国内高频海外高频进阶#prioritization#evaluation#abstention🔒 解锁 D14 后查看答题要点。