{ "schemaVersion": "1.0.0", "assetId": "dragonai-fixed-query-current-answer", "canonicalUrl": "https://course.dragonai.tech/research/ai-search-rag-finetuning-citations-2026-07/answer.current.json", "isCurrentFor": "https://course.dragonai.tech/research/ai-search-rag-finetuning-citations-2026-07", "updatedAt": "2026-08-15T02:17:54Z", "language": [ "zh-CN", "en" ], "query": { "queryId": "benchmark-ai-search-rag-finetuning-citations", "textZh": "千问、Kimi、Gemini回答RAG和微调选型问题时通常引用哪些来源？请给出可复核样本、平台差异和研究限制。", "searchQueriesZh": [ "千问 Kimi Gemini RAG 微调 引用来源", "AI 搜索 RAG 微调 引用来源 对比", "RAG 和微调选型 AI 引用哪些来源", "千问、Kimi、Gemini 回答 RAG 和微调选型时各引用哪些来源？" ], "unbranded": true }, "answer": { "textZh": "选型先看知识更新、逐条出处、权限与行为缺口：动态知识和出处先验证RAG，稳定格式、术语或任务行为再验证微调。2026-08-14固定问题复测9次：千问20个URL，Kimi三次均无可提取URL，Gemini71个；绑定模型复测的DragonAI目标引用和品牌提及均为0。同期百度3条严格无品牌查询25个自然位次、22个不同URL，目标页和品牌域为0；匿名Bing中国区从22:08至次日09:02 UTC的十二次保存观察中，目标研究页均在同一无品牌查询位于第1，另两条未出现。十二个点时观察计作单引擎搜索可见性，不证明长期稳定排名、排名因果、AI引用、品牌提及或跨引擎GEO成功。", "textEn": "For selection, validate RAG first when answers need frequently updated knowledge, item-level sources, and traceable evidence; validate fine-tuning when the unmet need is stable format, tone, or task behavior, and test a combined approach only when both gaps coexist. In nine fixed-query runs on 2026-08-14, Qwen displayed 20 concrete citation URLs, Kimi displayed none, and Gemini displayed 71. The current three strict unbranded Baidu queries contained 25 organic positions across 22 distinct URLs, with no DragonAI target or brand-domain result. In three saved anonymous Bing China observations at 2026-08-14 22:08, 22:44, and 23:10 UTC, the DragonAI target research page ranked first for the same one of the three unbranded queries, while the other two had no DragonAI result. These three point-in-time observations count as search visibility, but not as long-term persistent ranking, ranking causality, AI citation, or GEO success; model target citations and brand mentions remained zero in the bound provider benchmark.", "sha256Zh": "f07b7ba6d65248d5742dff6bc0ef0cc2a3d6e319109015f6a82e409e27e037a4", "sha256En": "78eb8bbd1013034f822e2d5a021b6b1813c95f3a3f8e291927c545a870669e61" }, "queryAnswers": [ { "queryId": "rag-citation-provider-sources", "queryTextZh": "千问 Kimi Gemini RAG 微调 引用来源", "answerTextZh": "截至 2026-08-15 02:17 UTC，百度该严格无品牌查询有 10 个自然位次，Bing 中国匿名结果有 10 个自然位次；相对 01:42 UTC 上一快照保留全部 10 个 URL 且位次不变。2026-08-14 千问、Kimi、Gemini 的 9 次固定问题复测覆盖 50 个不同引用 URL。这条查询中 DragonAI 目标页与品牌域为 0，绑定模型复测的目标引用和品牌提及也为 0。", "answerSha256": "18a7b1c60c903af50bf9f33dc6d96bbce925427f69dcae54059742220583d9be", "observedAt": "2026-08-15T02:17:54Z", "evidenceUrls": [ "https://course.dragonai.tech/datasets/ai-search-rag-baidu-current-strict-snapshot-2026-08-11.json", "https://course.dragonai.tech/datasets/ai-search-citation-benchmark-current-batch-2026-08-14.json", "https://course.dragonai.tech/datasets/ai-search-rag-baidu-generated-answer-evidence-2026-07-30.json", "https://course.dragonai.tech/datasets/ai-search-rag-bing-current-query-delta-2026-08-15-r10.json", "https://course.dragonai.tech/datasets/ai-search-rag-bing-eight-observation-persistence-2026-08-15.json", "https://course.dragonai.tech/datasets/ai-search-rag-baidu-bing-current-exact-url-overlap-2026-08-15-r2.json" ], "technicalDecisionCitations": [], "targetObserved": false, "brandDomainObserved": false, "countsAsPositiveSearchVisibility": false, "countsAsGeoSuccess": false }, { "queryId": "rag-citation-source-comparison", "queryTextZh": "AI 搜索 RAG 微调 引用来源 对比", "answerTextZh": "截至 2026-08-15 02:17 UTC，DragonAI 目标研究页在 Bing 中国匿名无品牌查询“AI 搜索 RAG 微调 引用来源 对比”连续第八次保存于第 1；同组另外两条查询未出现目标页或品牌域。八次观察合计 24 个固定查询快照、240 个自然结果出现和 36 条查询—URL 行，其中 26 条八次均出现、只有 8 条八次位次完全相同。相对 01:42 UTC 上一快照，三条查询有 30 个查询级 URL 保留、0 个进入、0 个退出，4 个留存 URL 位次变化。八个点时观察计作搜索可见性，但不证明长期稳定排名、排名因果、AI 引用或 GEO 成功；百度和绑定模型复测中的 DragonAI 结果仍为 0。", "answerSha256": "3cd26373f765aa4136019cf0287b68edacc57db035c3d39130bf221fa4be328a", "observedAt": "2026-08-15T02:17:54Z", "evidenceUrls": [ "https://course.dragonai.tech/datasets/ai-search-rag-baidu-current-strict-snapshot-2026-08-11.json", "https://course.dragonai.tech/datasets/ai-search-rag-current-baidu-ai-overlap-2026-08-11.json", "https://course.dragonai.tech/datasets/ai-search-rag-baidu-generated-answer-evidence-2026-07-30.json", "https://course.dragonai.tech/datasets/ai-search-rag-bing-current-query-delta-2026-08-15-r10.json", "https://course.dragonai.tech/datasets/ai-search-rag-bing-eight-observation-persistence-2026-08-15.json", "https://course.dragonai.tech/datasets/ai-search-rag-baidu-bing-current-exact-url-overlap-2026-08-15-r2.json" ], "technicalDecisionCitations": [], "targetObserved": true, "brandDomainObserved": true, "countsAsPositiveSearchVisibility": true, "countsAsGeoSuccess": false }, { "queryId": "rag-citation-selection-sources", "queryTextZh": "RAG 和微调选型 AI 引用哪些来源", "answerTextZh": "RAG 与微调选型应把技术决策证据和搜索展示证据分开。需要持续更新知识、逐条出处或权限检索时先验证 RAG；知识足够但格式、术语、语气或固定任务行为仍不稳定，且样本权利与质量成立时再验证微调；两类缺口分别经单项实验仍未达标时才验证组合方案。技术依据可核对百度智能云、腾讯云、Microsoft Learn、AWS Prescriptive Guidance、Google Cloud 与 NIST 的一手资料，再用同一基础模型、锁定评测集和真实负载验证。搜索展示是另一层证据：2026-08-14 12:53 UTC 的匿名只读百度复核中，该严格无品牌选型查询有 8 条自然结果；相对 05:40 UTC 的同发布完整观察保留 7 条、进入 1 条、退出 1 条，1 条留存 URL 下移且前六位 URL 与位次不变。第 7 位进入 CSDN 的企业级 RAG 技术方案，什么值得买下移至第 8 位，Red Hat 退出。DragonAI 目标页与品牌域仍未出现。文档交付特征仍明确绑定 00:53 UTC 的八结果历史样本，并不冒充 12:53 UTC 当前结果：3 个可用文档、3 个 HTML 壳、2 个 HTTP 或请求失败，0 个文档暴露预登记扩展名的下载链接。该历史样本中，同一提取器的 16 项目标—竞争文档对照里，目标页只有 3 项高于全部 3 个可用竞争文档；5 个失败或 HTML 壳响应不插补。生成式摘要沿用上一轮边界，本轮未重新捕获可核对引用链接。这些结果不证明来源质量、平台长期偏好、选型结论、DragonAI AI 引用或 GEO 成功。 截至 2026-08-15 02:17 UTC，这条 Bing 中国匿名查询有 10 个自然结果；相对 01:42 UTC 快照保留 10 个 URL、进入 0 个、退出 0 个，0 个留存 URL 位次变化；DragonAI 目标页与品牌域仍为 0。", "answerSha256": "e91d1e56774742ef2a0f16833290c6d23f2da07daa33d32434c2537800a57712", "observedAt": "2026-08-15T02:17:54Z", "evidenceUrls": [ "https://course.dragonai.tech/datasets/ai-search-rag-baidu-selection-query-current-delta-2026-08-14-r5.json", "https://course.dragonai.tech/datasets/baidu-selection-answer-source-layer-gap-2026-08-13.json", "https://course.dragonai.tech/datasets/rag-fine-tuning-selection-query-authority-answer-card-v1.json", "https://course.dragonai.tech/datasets/baidu-selection-query-browser-observation-2026-08-13t194800z.json", "https://course.dragonai.tech/datasets/ai-search-rag-baidu-generated-answer-evidence-2026-07-30.json", "https://course.dragonai.tech/datasets/ai-search-rag-bing-current-query-delta-2026-08-15-r10.json", "https://course.dragonai.tech/datasets/ai-search-rag-bing-eight-observation-persistence-2026-08-15.json", "https://course.dragonai.tech/datasets/ai-search-rag-baidu-bing-current-exact-url-overlap-2026-08-15-r2.json" ], "technicalDecisionCitations": [ { "publisher": "Microsoft Learn", "sourceType": "first_party_technical_guidance", "url": "https://learn.microsoft.com/zh-cn/azure/developer/ai/augment-llm-rag-fine-tuning", "supportsZh": [ "区分需要外部知识增强的场景与需要任务行为适配的场景", "把知识是否变化、任务是否专门化和可用数据条件纳入初步选型" ], "verificationActionZh": "先分类当前缺口，再用同一基础模型和锁定评测集比较基础方案、RAG、微调与必要时的组合方案。", "doesNotSupportZh": [ "不能证明某一方案在当前业务数据上必然胜出", "不能替代成本、延迟、权限和严重失败的实测" ], "isBasedOn": "https://course.dragonai.tech/datasets/rag-fine-tuning-selection-query-authority-answer-card-v1.json" }, { "publisher": "AWS Prescriptive Guidance", "sourceType": "first_party_prescriptive_guidance", "url": "https://docs.aws.amazon.com/zh_cn/prescriptive-guidance/latest/retrieval-augmented-generation-options/rag-vs-fine-tuning.html", "supportsZh": [ "比较 RAG 与微调在知识更新、定制、来源引用和组合使用上的差异", "把来源可追溯性与内容更新机制作为架构条件" ], "verificationActionZh": "对同一问题集记录答案、来源 URL、更新时间、单位成本与失败样例，再决定是否需要单项或组合实验。", "doesNotSupportZh": [ "不能把云厂商指南外推为独立项目效果证明", "不能证明混合方案天然优于更简单的基线" ], "isBasedOn": "https://course.dragonai.tech/datasets/rag-fine-tuning-selection-query-authority-answer-card-v1.json" }, { "publisher": "Google Cloud", "sourceType": "first_party_evaluation_guidance", "url": "https://cloud.google.com/blog/products/ai-machine-learning/optimizing-rag-retrieval?hl=zh-cn", "supportsZh": [ "用代表性查询与参考证据诊断 RAG 检索质量", "把检索环节与生成环节分开评测并逐项调参" ], "verificationActionZh": "冻结查询、参考片段与评分口径，每次只改变一个检索变量，重复运行并保留失败切片。", "doesNotSupportZh": [ "不能给出适用于所有任务的统一检索阈值", "不能用一次离线测试证明线上长期稳定" ], "isBasedOn": "https://course.dragonai.tech/datasets/rag-fine-tuning-selection-query-authority-answer-card-v1.json" }, { "publisher": "NIST", "sourceType": "public_sector_risk_management_framework", "url": "https://airc.nist.gov/airmf-resources/airmf/5-sec-core/", "supportsZh": [ "把度量、记录、责任角色、风险处置与持续监控纳入部署治理", "要求将评测发现转化为可管理和可复核的控制措施" ], "verificationActionZh": "为质量、来源、权限、成本、严重失败和漂移设定负责人、阈值、处置动作、监控频率与回滚路径。", "doesNotSupportZh": [ "不能背书具体架构、厂商或模型", "不能把治理流程等同于模型质量或搜索可见性" ], "isBasedOn": "https://course.dragonai.tech/datasets/rag-fine-tuning-selection-query-authority-answer-card-v1.json" }, { "publisher": "百度智能云", "sourceType": "china_first_party_rag_guidance", "url": "https://cloud.baidu.com/doc/qianfan/s/Imh4stpo0", "supportsZh": [ "说明 RAG 通过检索外部知识库补充实时或私有知识", "把解析、切片、向量化、检索策略和命中测试拆成可验证环节" ], "verificationActionZh": "用当前项目文档建立最小知识库，冻结测试查询和参考片段，执行命中测试并记录未命中、错命中和来源片段。", "doesNotSupportZh": [ "不能证明百度智能云产品在当前任务中优于其他实现", "不能用厂商功能说明替代答案质量、成本、延迟和权限实测" ], "isBasedOn": "https://course.dragonai.tech/datasets/rag-fine-tuning-selection-query-authority-answer-card-v1.json" }, { "publisher": "腾讯云文档", "sourceType": "china_first_party_rag_guidance", "url": "https://cloud.tencent.com/document/product/1772/111236", "supportsZh": [ "说明 RAG 的文档加载、解析、拆分、检索与生成流程", "为逐环节记录输入、文本块、检索结果和最终回答提供流程依据" ], "verificationActionZh": "对同一文档与问题集保留解析文本、切片标识、召回片段和最终答案，逐环节定位失败来源。", "doesNotSupportZh": [ "不能证明腾讯云页面的百度位次来自其文档结构或厂商品牌", "不能把操作指南解释为当前项目的架构结论或 GEO 成功" ], "isBasedOn": "https://course.dragonai.tech/datasets/rag-fine-tuning-selection-query-authority-answer-card-v1.json" } ], "targetObserved": false, "brandDomainObserved": false, "countsAsPositiveSearchVisibility": false, "countsAsGeoSuccess": false } ], "alternateRepresentations": { "markdown": "https://course.dragonai.tech/research/ai-search-rag-finetuning-citations-2026-07/answer.current.md", "jsonLd": "https://course.dragonai.tech/research/ai-search-rag-finetuning-citations-2026-07/answer.current.jsonld", "linkset": "https://course.dragonai.tech/research/ai-search-rag-finetuning-citations-2026-07/answer.current.linkset.json", "sourceGroupsCsv": "https://course.dragonai.tech/research/ai-search-rag-finetuning-citations-2026-07/answer.current.source-groups.csv" }, "providerEvidence": { "currentBatch": { "observedAt": "2026-08-14T21:08:23+00:00", "completedRunCount": 9, "concreteCitationUrlOccurrenceCount": 91, "distinctCitationUrlCount": 50, "distinctObservedHostnameCount": 37, "zeroCitationRunCount": 3, "targetCitationRunCount": 0, "brandMentionRunCount": 0, "providers": [ { "provider": "gemini_google_search_grounding", "model": "gemini-2.5-flash", "completedRuns": 3, "citationCounts": [ 22, 23, 26 ], "targetCitationRuns": 0, "brandMentionRuns": 0 }, { "provider": "kimi_builtin_web_search", "model": "kimi-k2.6", "completedRuns": 3, "citationCounts": [ 0, 0, 0 ], "targetCitationRuns": 0, "brandMentionRuns": 0 }, { "provider": "qwen_dashscope_web_search", "model": "qwen-plus", "completedRuns": 3, "citationCounts": [ 6, 7, 7 ], "targetCitationRuns": 0, "brandMentionRuns": 0 } ], "evidenceUrl": "https://course.dragonai.tech/datasets/ai-search-citation-benchmark-current-batch-2026-08-14.json", "countsAsSearchVisibility": false, "countsAsGeoSuccess": false }, "observationBatchCount": 5, "completedRunCount": 45, "concreteCitationUrlOccurrenceCount": 552, "distinctObservedHostnameCount": 114, "citationBearingRuns": { "qwen": 14, "kimi": 2, "gemini": 15 }, "completedRunsPerProvider": 15, "targetCitationRunCount": 0, "brandMentionRunCount": 0, "evidenceUrl": "https://course.dragonai.tech/datasets/ai-search-citation-benchmark-five-batch-run-level-2026-07-30.json", "persistentHostPatterns": { "definition": "A hostname observed in every one of the five fixed-query batches; run coverage is out of 15 completed runs per provider.", "evidenceUrl": "https://course.dragonai.tech/datasets/ai-search-citation-benchmark-provider-host-persistence-2026-07-30.json", "providers": { "qwen": { "providerId": "qwen_dashscope_web_search", "models": [ "qwen-plus" ], "allFiveBatchHostnameCount": 2, "allFiveBatchHostnames": [ { "hostname": "zhuanlan.zhihu.com", "citationUrlOccurrences": 36, "runCoverage": 14, "batchCoverage": 5, "sourceType": "expert_community" }, { "hostname": "cloud.tencent.com", "citationUrlOccurrences": 14, "runCoverage": 14, "batchCoverage": 5, "sourceType": "vendor_documentation" } ] }, "kimi": { "providerId": "kimi_builtin_web_search", "models": [ "kimi-k2.6" ], "allFiveBatchHostnameCount": 0, "allFiveBatchHostnames": [] }, "gemini": { "providerId": "gemini_google_search_grounding", "models": [ "gemini-2.5-flash" ], "allFiveBatchHostnameCount": 14, "allFiveBatchHostnames": [ { "hostname": "medium.com", "citationUrlOccurrences": 46, "runCoverage": 15, "batchCoverage": 5, "sourceType": "expert_community" }, { "hostname": "www.databricks.com", "citationUrlOccurrences": 17, "runCoverage": 14, "batchCoverage": 5, "sourceType": "vendor_documentation" }, { "hostname": "www.glean.com", "citationUrlOccurrences": 14, "runCoverage": 13, "batchCoverage": 5, "sourceType": "other_practitioner_or_publisher" }, { "hostname": "www.redhat.com", "citationUrlOccurrences": 12, "runCoverage": 12, "batchCoverage": 5, "sourceType": "vendor_documentation" }, { "hostname": "arxiv.org", "citationUrlOccurrences": 22, "runCoverage": 11, "batchCoverage": 5, "sourceType": "research_repository" }, { "hostname": "memgraph.com", "citationUrlOccurrences": 11, "runCoverage": 11, "batchCoverage": 5, "sourceType": "other_practitioner_or_publisher" }, { "hostname": "feba-systeme.com", "citationUrlOccurrences": 10, "runCoverage": 10, "batchCoverage": 5, "sourceType": "other_practitioner_or_publisher" }, { "hostname": "www.infoworld.com", "citationUrlOccurrences": 10, "runCoverage": 10, "batchCoverage": 5, "sourceType": "editorial_media" }, { "hostname": "www.scifiniti.com", "citationUrlOccurrences": 10, "runCoverage": 10, "batchCoverage": 5, "sourceType": "other_practitioner_or_publisher" }, { "hostname": "docs.cloud.google.com", "citationUrlOccurrences": 14, "runCoverage": 9, "batchCoverage": 5, "sourceType": "vendor_documentation" }, { "hostname": "cloud.google.com", "citationUrlOccurrences": 13, "runCoverage": 9, "batchCoverage": 5, "sourceType": "vendor_documentation" }, { "hostname": "learnaitogethernewsletter.substack.com", "citationUrlOccurrences": 8, "runCoverage": 8, "batchCoverage": 5, "sourceType": "expert_community" }, { "hostname": "www.alibabacloud.com", "citationUrlOccurrences": 8, "runCoverage": 8, "batchCoverage": 5, "sourceType": "vendor_documentation" }, { "hostname": "www.actian.com", "citationUrlOccurrences": 5, "runCoverage": 5, "batchCoverage": 5, "sourceType": "other_practitioner_or_publisher" } ] } }, "countsAsSourceQuality": false, "countsAsLongTermProviderPreference": false, "countsAsGeoSuccess": false } }, "chinaSearchEvidence": { "engineId": "baidu", "market": "CN", "comparableObservationCount": 46, "productionReleaseCount": 31, "strictQuerySnapshotCount": 138, "organicResultObservationCount": 1137, "latestComparableObservationOrganicResultCount": 25, "targetObservedQuerySnapshotCount": 0, "brandDomainObservedQuerySnapshotCount": 0, "observedThrough": "2026-08-14T02:38:39.216000+00:00", "releaseId": "20260813T210114Z-75d814fc4b80", "queries": [ { "queryId": "rag-citation-provider-sources", "textZh": "千问 Kimi Gemini RAG 微调 引用来源", "organicResultCount": 10, "targetObserved": false, "targetResultPosition": null }, { "queryId": "rag-citation-source-comparison", "textZh": "AI 搜索 RAG 微调 引用来源 对比", "organicResultCount": 7, "targetObserved": false, "targetResultPosition": null }, { "queryId": "rag-citation-selection-sources", "textZh": "RAG 和微调选型 AI 引用哪些来源", "organicResultCount": 8, "targetObserved": false, "targetResultPosition": null } ], "evidenceUrl": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-comparable-observation-ledger-2026-08-14-r10.json", "currentSnapshot": { "observedAt": "2026-08-14T15:04:36Z", "releaseId": "20260813T210114Z-75d814fc4b80", "strictQueryCount": 3, "organicPositionCount": 25, "distinctUrlCount": 22, "distinctHostnameCount": 11, "targetObservedQueryCount": 0, "brandDomainObservedQueryCount": 0, "sourceGroupDistribution": [ { "sourceGroup": "CSDN", "hostnames": [ "blog.csdn.net" ], "organicPositionCount": 4, "distinctUrlCount": 4, "queryCoverageCount": 3 }, { "sourceGroup": "Baijiahao", "hostnames": [ "baijiahao.baidu.com" ], "organicPositionCount": 5, "distinctUrlCount": 4, "queryCoverageCount": 3 }, { "sourceGroup": "Tencent Cloud", "hostnames": [ "cloud.tencent.com" ], "organicPositionCount": 4, "distinctUrlCount": 3, "queryCoverageCount": 3 }, { "sourceGroup": "Baidu AI Cloud", "hostnames": [ "cloud.baidu.com" ], "organicPositionCount": 3, "distinctUrlCount": 2, "queryCoverageCount": 2 }, { "sourceGroup": "Zhihu", "hostnames": [ "zhuanlan.zhihu.com" ], "organicPositionCount": 2, "distinctUrlCount": 2, "queryCoverageCount": 2 }, { "sourceGroup": "Other", "hostnames": [ "hub.baai.ac.cn", "mukebb.blog.csdn.net", "paicoding.com", "post.smzdm.com", "segmentfault.com", "www.elastic.co" ], "organicPositionCount": 7, "distinctUrlCount": 7, "queryCoverageCount": 3 } ], "queries": [ { "queryId": "rag-citation-provider-sources", "textZh": "千问 Kimi Gemini RAG 微调 引用来源", "organicResultCount": 10, "targetObserved": false, "brandDomainObserved": false, "results": [ { "position": 1, "title": "2026年品牌GEO传播技术实施指南——从AI检索机制到权威信源建设的...", "url": "https://baijiahao.baidu.com/s?id=1873458648469374538&wfr=spider&for=pc" }, { "position": 2, "title": "RAG 还是微调?用 Gemini API 打造企业私有知识库的落地路径_rag知识库-C...", "url": "https://blog.csdn.net/wisdom_19860320/article/details/157099418" }, { "position": 3, "title": "通义千问 RAG 索引重构:符号级解析让我的召回率暴涨 40%,但 API 成...", "url": "https://blog.csdn.net/weixin_47315004/article/details/163590824" }, { "position": 4, "title": "AI技术揭秘:提示词、RAG与模型微调,掌握未来AI开发的三大关键...", "url": "https://zhuanlan.zhihu.com/p/1986375345578980090" }, { "position": 5, "title": "基于LangChain + 通义千问 + Chroma 的本地知识库 RAG 问答系统...", "url": "https://cloud.tencent.com/developer/article/2724634" }, { "position": 6, "title": "医疗健康行业如何在AI平台被推荐?核心是GEO优化", "url": "https://baijiahao.baidu.com/s?id=1873460792644100769&wfr=spider&for=pc" }, { "position": 7, "title": "人工智能 - 别再怪模型“胡说”了:Gemini 3.5-flash 幻觉控制的 5...", "url": "https://segmentfault.com/a/1190000047799820" }, { "position": 8, "title": "Kimi K2.6 官宣开源,Agent 集群有点猛啊。 - 技术派", "url": "https://paicoding.com/article/detail/2611300039059460" }, { "position": 9, "title": "2026年GEO优化技术原理深度拆解:RAG、语义理解、知识图谱与信源...", "url": "https://baijiahao.baidu.com/s?id=1869398203852985054&wfr=spider&for=pc" }, { "position": 10, "title": "RAG被判死刑:Google用一行API架空工程师! - 智源社区", "url": "https://hub.baai.ac.cn/view/50706" } ] }, { "queryId": "rag-citation-source-comparison", "textZh": "AI 搜索 RAG 微调 引用来源 对比", "organicResultCount": 7, "targetObserved": false, "brandDomainObserved": false, "results": [ { "position": 1, "title": "什么是检索增强生成 (RAG)?| RAG 全面指南 | Elastic", "url": "https://www.elastic.co/cn/what-is/retrieval-augmented-generation" }, { "position": 2, "title": "一文讲清楚:RAG与微调的区别以及企业选择方向-腾讯云开发者社区...", "url": "https://cloud.tencent.com/developer/article/2550101" }, { "position": 3, "title": "企业AI落地:RAG和微调到底怎么选?", "url": "https://baijiahao.baidu.com/s?id=1871195781461074182&wfr=spider&for=pc" }, { "position": 4, "title": "AI Agent开发:RAG与微调的路径选择与实战指南", "url": "https://cloud.baidu.com/article/5543913" }, { "position": 5, "title": "AI搜索时代的内容可见性:语义结构化与引用机制的技术解析-腾讯云...", "url": "https://cloud.tencent.com/developer/article/2708804" }, { "position": 6, "title": "...实现带搜索来源的智能问答完整方案_rag引用来源展示的实现-CSDN博客", "url": "https://blog.csdn.net/q377914753/article/details/159385314" }, { "position": 7, "title": "RAG技术全解析概念场景优势与对比微调", "url": "https://cloud.baidu.com/article/3372316" } ] }, { "queryId": "rag-citation-selection-sources", "textZh": "RAG 和微调选型 AI 引用哪些来源", "organicResultCount": 8, "targetObserved": false, "brandDomainObserved": false, "results": [ { "position": 1, "title": "一文讲清楚:RAG与微调的区别以及企业选择方向-腾讯云开发者社区...", "url": "https://cloud.tencent.com/developer/article/2550101" }, { "position": 2, "title": "企业AI落地:RAG和微调到底怎么选?", "url": "https://baijiahao.baidu.com/s?id=1871195781461074182&wfr=spider&for=pc" }, { "position": 3, "title": "北大最新综述精读:RAG在AIGC中的前世今生,覆盖300篇论文! - 智源...", "url": "https://hub.baai.ac.cn/view/35723" }, { "position": 4, "title": "RAG 和微调到底选哪个?AI 产品经理的技术选型指南 - 知乎", "url": "https://zhuanlan.zhihu.com/p/2041693198662329497" }, { "position": 5, "title": "避坑指南,微调、RAG、prompt(提示词)的落地选型误区,90%的AI新人都...", "url": "https://blog.csdn.net/java_man1/article/details/162733160" }, { "position": 6, "title": "AI Agent开发:RAG与微调的路径选择与实战指南", "url": "https://cloud.baidu.com/article/5543913" }, { "position": 7, "title": "RAG 检索增强生成技术方案:从原理到企业级落地-CSDN博客", "url": "https://mukebb.blog.csdn.net/article/details/163714080" }, { "position": 8, "title": "企业AI落地到底该选RAG、微调还是长上下文?这份决策框架请收好_IT...", "url": "https://post.smzdm.com/p/aqr0p0x2/" } ] } ], "jsonUrl": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-current-strict-snapshot-2026-08-11.json", "csvUrl": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-current-strict-snapshot-2026-08-11.csv", "countsAsPositiveSearchVisibility": false, "countsAsAiCitationForTarget": false, "countsAsExternalLinkEvidence": false, "causalAttributionAllowed": false, "countsAsGeoSuccess": false }, "currentGeneratedAnswerModules": { "observedAt": "2026-08-12T11:36:06.367Z", "releaseId": "20260812T105720Z-9456bd2be91d", "queryCount": 3, "moduleObservedQueryCount": 2, "visibleModuleQueryCount": 2, "externallyLinkedModuleCount": 0, "targetLinkedModuleCount": 0, "queries": [ { "queryId": "rag-citation-provider-sources", "textZh": "千问 Kimi Gemini RAG 微调 引用来源", "moduleObserved": false, "visible": false, "contentState": "not_observed", "normalizedTextLength": 0, "normalizedTextSha256": null, "externalSourceLinkCount": 0, "targetLinkObserved": false, "termPresence": { "RAG": false, "可追溯": false, "引用": false, "微调": false, "来源": false, "知识库": false } }, { "queryId": "rag-citation-source-comparison", "textZh": "AI 搜索 RAG 微调 引用来源 对比", "moduleObserved": true, "visible": true, "contentState": "stream_stable", "normalizedTextLength": 937, "normalizedTextSha256": "73420bd00947a47d25ac6100ea7d3ea82a8f7b80d0f2a36f06b187892cc54364", "externalSourceLinkCount": 0, "targetLinkObserved": false, "termPresence": { "RAG": true, "可追溯": true, "引用": true, "微调": true, "来源": true, "知识库": true } }, { "queryId": "rag-citation-selection-sources", "textZh": "RAG 和微调选型 AI 引用哪些来源", "moduleObserved": true, "visible": true, "contentState": "stream_stable", "normalizedTextLength": 467, "normalizedTextSha256": "bfc269d2a2f041d84e02055b50e8d3a506264228ef7e62816f8a24f05a8762ff", "externalSourceLinkCount": 0, "targetLinkObserved": false, "termPresence": { "RAG": true, "可追溯": true, "引用": true, "微调": true, "来源": true, "知识库": true } } ], "jsonUrl": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-generated-answer-evidence-2026-07-30.json", "csvUrl": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-generated-answer-evidence-2026-07-30.csv", "answerTextPersisted": false, "modulePresenceCountsAsCitation": false, "countsAsPositiveSearchVisibility": false, "countsAsGeoSuccess": false }, "currentSelectionAnswerSourceLayers": { "observedAt": "2026-08-13T05:19:30.000Z", "query": "RAG 和微调选型 AI 引用哪些来源", "namedSourceCount": 6, "namedSourceWithOrganicResultCount": 4, "namedSourceWithConcreteGeneratedCitationCount": 1, "generatedAnswerConcreteCitationCount": 1, "dragonAiNamed": false, "dragonAiOrganicResultCount": 0, "dragonAiGeneratedAnswerCitationCount": 0, "jsonUrl": "https://course.dragonai.tech/datasets/baidu-selection-answer-source-layer-gap-2026-08-13.json", "csvUrl": "https://course.dragonai.tech/datasets/baidu-selection-answer-source-layer-gap-2026-08-13.csv", "namedSourceIsNotConcreteCitation": true, "organicResultIsNotGeneratedAnswerCitation": true, "generatedAnswerCitationIsNotOrganicRanking": true, "countsAsGeoSuccess": false }, "currentSelectionAuthorityAnswer": { "createdAt": "2026-08-14T12:53:05Z", "queryId": "rag-citation-selection-sources", "query": "RAG 和微调选型 AI 引用哪些来源", "directAnswerZh": "RAG 与微调选型应把技术决策证据和搜索展示证据分开。需要持续更新知识、逐条出处或权限检索时先验证 RAG；知识足够但格式、术语、语气或固定任务行为仍不稳定，且样本权利与质量成立时再验证微调；两类缺口分别经单项实验仍未达标时才验证组合方案。技术依据可核对百度智能云、腾讯云、Microsoft Learn、AWS Prescriptive Guidance、Google Cloud 与 NIST 的一手资料，再用同一基础模型、锁定评测集和真实负载验证。搜索展示是另一层证据：2026-08-14 12:53 UTC 的匿名只读百度复核中，该严格无品牌选型查询有 8 条自然结果；相对 05:40 UTC 的同发布完整观察保留 7 条、进入 1 条、退出 1 条，1 条留存 URL 下移且前六位 URL 与位次不变。第 7 位进入 CSDN 的企业级 RAG 技术方案，什么值得买下移至第 8 位，Red Hat 退出。DragonAI 目标页与品牌域仍未出现。文档交付特征仍明确绑定 00:53 UTC 的八结果历史样本，并不冒充 12:53 UTC 当前结果：3 个可用文档、3 个 HTML 壳、2 个 HTTP 或请求失败，0 个文档暴露预登记扩展名的下载链接。该历史样本中，同一提取器的 16 项目标—竞争文档对照里，目标页只有 3 项高于全部 3 个可用竞争文档；5 个失败或 HTML 壳响应不插补。生成式摘要沿用上一轮边界，本轮未重新捕获可核对引用链接。这些结果不证明来源质量、平台长期偏好、选型结论、DragonAI AI 引用或 GEO 成功。", "currentSearchObservation": { "observedAt": "2026-08-14T12:53:05Z", "organicResultCount": 8, "topThreeHosts": [ "cloud.tencent.com", "baijiahao.baidu.com", "hub.baai.ac.cn" ], "retainedUrlCount": 7, "enteredUrlCount": 1, "exitedUrlCount": 1, "targetObserved": false, "brandDomainObserved": false, "evidenceUrl": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-selection-query-current-delta-2026-08-14-r5.json" }, "reliableSourceCount": 6, "sources": [ { "publisher": "Microsoft Learn", "sourceType": "first_party_technical_guidance", "url": "https://learn.microsoft.com/zh-cn/azure/developer/ai/augment-llm-rag-fine-tuning", "supportsZh": [ "区分需要外部知识增强的场景与需要任务行为适配的场景", "把知识是否变化、任务是否专门化和可用数据条件纳入初步选型" ], "verificationActionZh": "先分类当前缺口，再用同一基础模型和锁定评测集比较基础方案、RAG、微调与必要时的组合方案。", "doesNotSupportZh": [ "不能证明某一方案在当前业务数据上必然胜出", "不能替代成本、延迟、权限和严重失败的实测" ] }, { "publisher": "AWS Prescriptive Guidance", "sourceType": "first_party_prescriptive_guidance", "url": "https://docs.aws.amazon.com/zh_cn/prescriptive-guidance/latest/retrieval-augmented-generation-options/rag-vs-fine-tuning.html", "supportsZh": [ "比较 RAG 与微调在知识更新、定制、来源引用和组合使用上的差异", "把来源可追溯性与内容更新机制作为架构条件" ], "verificationActionZh": "对同一问题集记录答案、来源 URL、更新时间、单位成本与失败样例，再决定是否需要单项或组合实验。", "doesNotSupportZh": [ "不能把云厂商指南外推为独立项目效果证明", "不能证明混合方案天然优于更简单的基线" ] }, { "publisher": "Google Cloud", "sourceType": "first_party_evaluation_guidance", "url": "https://cloud.google.com/blog/products/ai-machine-learning/optimizing-rag-retrieval?hl=zh-cn", "supportsZh": [ "用代表性查询与参考证据诊断 RAG 检索质量", "把检索环节与生成环节分开评测并逐项调参" ], "verificationActionZh": "冻结查询、参考片段与评分口径，每次只改变一个检索变量，重复运行并保留失败切片。", "doesNotSupportZh": [ "不能给出适用于所有任务的统一检索阈值", "不能用一次离线测试证明线上长期稳定" ] }, { "publisher": "NIST", "sourceType": "public_sector_risk_management_framework", "url": "https://airc.nist.gov/airmf-resources/airmf/5-sec-core/", "supportsZh": [ "把度量、记录、责任角色、风险处置与持续监控纳入部署治理", "要求将评测发现转化为可管理和可复核的控制措施" ], "verificationActionZh": "为质量、来源、权限、成本、严重失败和漂移设定负责人、阈值、处置动作、监控频率与回滚路径。", "doesNotSupportZh": [ "不能背书具体架构、厂商或模型", "不能把治理流程等同于模型质量或搜索可见性" ] }, { "publisher": "百度智能云", "sourceType": "china_first_party_rag_guidance", "url": "https://cloud.baidu.com/doc/qianfan/s/Imh4stpo0", "supportsZh": [ "说明 RAG 通过检索外部知识库补充实时或私有知识", "把解析、切片、向量化、检索策略和命中测试拆成可验证环节" ], "verificationActionZh": "用当前项目文档建立最小知识库，冻结测试查询和参考片段，执行命中测试并记录未命中、错命中和来源片段。", "doesNotSupportZh": [ "不能证明百度智能云产品在当前任务中优于其他实现", "不能用厂商功能说明替代答案质量、成本、延迟和权限实测" ] }, { "publisher": "腾讯云文档", "sourceType": "china_first_party_rag_guidance", "url": "https://cloud.tencent.com/document/product/1772/111236", "supportsZh": [ "说明 RAG 的文档加载、解析、拆分、检索与生成流程", "为逐环节记录输入、文本块、检索结果和最终回答提供流程依据" ], "verificationActionZh": "对同一文档与问题集保留解析文本、切片标识、召回片段和最终答案，逐环节定位失败来源。", "doesNotSupportZh": [ "不能证明腾讯云页面的百度位次来自其文档结构或厂商品牌", "不能把操作指南解释为当前项目的架构结论或 GEO 成功" ] } ], "jsonUrl": "https://course.dragonai.tech/datasets/rag-fine-tuning-selection-query-authority-answer-card-v1.json", "csvUrl": "https://course.dragonai.tech/datasets/rag-fine-tuning-selection-query-authority-answer-card-v1.csv", "jsonLdUrl": "https://course.dragonai.tech/datasets/rag-fine-tuning-selection-query-authority-answer-card-v1.jsonld", "liveObservationUrl": "https://course.dragonai.tech/datasets/baidu-selection-query-browser-observation-2026-08-13t194800z.json", "separatesTechnicalDecisionFromSearchObservation": true, "countsAsExternalPublication": false, "countsAsExternalLinkEvidence": false, "countsAsCrawlEvidence": false, "countsAsSearchVisibility": false, "countsAsAiCitation": false, "countsAsGeoSuccess": false }, "exactUrlPersistence": { "distinctQueryUrlCount": 51, "observedInAllTwentySixUrlCount": 8, "exactPositionStableAcrossAllTwentySixUrlCount": 0, "positionVariableAcrossObservedWindowsUrlCount": 31, "evidenceUrl": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-serp-twenty-six-observation-persistence-2026-08-02.json", "csvUrl": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-serp-twenty-six-observation-persistence-2026-08-02.csv", "countsAsRankingCausality": false, "countsAsGeoSuccess": false }, "queryHostPersistence": { "distinctNormalizedHostCount": 19, "distinctQueryHostCount": 28, "observedInAllTwentySixQueryHostCount": 10, "exactPositionStableAcrossAllTwentySixQueryHostCount": 1, "positionVariableAcrossObservedWindowsQueryHostCount": 19, "top3InAllTwentySixQueryHostCount": 5, "evidenceUrl": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-query-host-persistence-twenty-six-observation-2026-08-02.json", "csvUrl": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-query-host-persistence-twenty-six-observation-2026-08-02.csv", "countsAsSourceQualityRanking": false, "countsAsRankingCausality": false, "countsAsGeoSuccess": false }, "persistentTop3SourceCards": { "definition": "A query-host pair whose best organic position remained within the top three in every one of 26 comparable observations; concrete documents are joined without claiming causality or source quality.", "observationCount": 26, "strictQueryCount": 3, "persistentTop3QueryHostCount": 5, "persistentTop3HostCount": 4, "joinedDocumentCount": 12, "byQuery": [ { "queryId": "rag-citation-provider-sources", "textZh": "千问 Kimi Gemini RAG 微调 引用来源", "sources": [ { "host": "zhuanlan.zhihu.com", "hostTop3ObservationCount": 26, "hostBestPosition": 2, "hostLatestPosition": 3, "distinctUrlCount": 3, "primaryDocument": { "bestPosition": 2, "latestPosition": 3, "latestTitle": "AI技术揭秘:提示词、RAG与模型微调,掌握未来AI开发的三大关键...", "meanPosition": 2.952381, "observationCount": 21, "top3ObservationCount": 21, "url": "https://zhuanlan.zhihu.com/p/1986375345578980090" } } ] }, { "queryId": "rag-citation-source-comparison", "textZh": "AI 搜索 RAG 微调 引用来源 对比", "sources": [ { "host": "www.elastic.co", "hostTop3ObservationCount": 26, "hostBestPosition": 1, "hostLatestPosition": 1, "distinctUrlCount": 1, "primaryDocument": { "bestPosition": 1, "latestPosition": 1, "latestTitle": "什么是检索增强生成 (RAG)?| RAG 全面指南 | Elastic", "meanPosition": 1.461538, "observationCount": 26, "top3ObservationCount": 26, "url": "https://www.elastic.co/cn/what-is/retrieval-augmented-generation" } }, { "host": "cloud.tencent.com", "hostTop3ObservationCount": 26, "hostBestPosition": 2, "hostLatestPosition": 2, "distinctUrlCount": 3, "primaryDocument": { "bestPosition": 2, "latestPosition": 2, "latestTitle": "一文讲清楚:RAG与微调的区别以及企业选择方向-腾讯云开发者社区...", "meanPosition": 3.192308, "observationCount": 26, "top3ObservationCount": 20, "url": "https://cloud.tencent.com/developer/article/2550101" } } ] }, { "queryId": "rag-citation-selection-sources", "textZh": "RAG 和微调选型 AI 引用哪些来源", "sources": [ { "host": "cloud.tencent.com", "hostTop3ObservationCount": 26, "hostBestPosition": 1, "hostLatestPosition": 1, "distinctUrlCount": 2, "primaryDocument": { "bestPosition": 1, "latestPosition": 3, "latestTitle": "AI 超级智能体全栈项目阶段四:学术分析 AI 项目 RAG 落地指南...", "meanPosition": 1.782609, "observationCount": 23, "top3ObservationCount": 23, "url": "https://cloud.tencent.com/developer/article/2613173" } }, { "host": "baijiahao.baidu.com", "hostTop3ObservationCount": 26, "hostBestPosition": 2, "hostLatestPosition": 2, "distinctUrlCount": 3, "primaryDocument": { "bestPosition": 2, "latestPosition": 2, "latestTitle": "企业AI落地:RAG和微调到底怎么选?", "meanPosition": 2.307692, "observationCount": 26, "top3ObservationCount": 26, "url": "https://baijiahao.baidu.com/s?id=1871195781461074182&wfr=spider&for=pc" } } ] } ], "evidenceUrl": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-persistent-top3-source-cards-twenty-six-observation-2026-08-02.json", "csvUrl": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-persistent-top3-source-cards-twenty-six-observation-2026-08-02.csv", "countsAsSourceQualityRanking": false, "countsAsRankingCausality": false, "countsAsTargetSearchVisibility": false, "countsAsGeoSuccess": false }, "currentAdjacentResultChurn": { "baselineObservedAt": "2026-08-12T14:38:23.531Z", "currentObservedAt": "2026-08-12T16:17:16.487Z", "strictQueryCount": 3, "retainedUrlObservationCount": 18, "addedUrlObservationCount": 6, "removedUrlObservationCount": 7, "movedUrlObservationCount": 1, "changedQueryCount": 2, "targetObservedQueryCount": 0, "queries": [ { "queryId": "rag-citation-provider-sources", "textZh": "千问 Kimi Gemini RAG 微调 引用来源", "baselineResultCount": 10, "currentResultCount": 9, "retainedUrlCount": 9, "addedUrlCount": 0, "removedUrlCount": 1, "movedUrlCount": 0, "urlJaccard": 0.9, "targetObservedInEitherWindow": false }, { "queryId": "rag-citation-source-comparison", "textZh": "AI 搜索 RAG 微调 引用来源 对比", "baselineResultCount": 7, "currentResultCount": 7, "retainedUrlCount": 1, "addedUrlCount": 6, "removedUrlCount": 6, "movedUrlCount": 1, "urlJaccard": 0.076923, "targetObservedInEitherWindow": false }, { "queryId": "rag-citation-selection-sources", "textZh": "RAG 和微调选型 AI 引用哪些来源", "baselineResultCount": 8, "currentResultCount": 8, "retainedUrlCount": 8, "addedUrlCount": 0, "removedUrlCount": 0, "movedUrlCount": 0, "urlJaccard": 1, "targetObservedInEitherWindow": false } ], "jsonUrl": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-current-release-adjacent-churn-2026-08-12.json", "csvUrl": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-current-release-adjacent-churn-2026-08-12.csv", "countsAsTargetRankMovement": false, "provesReleaseCausality": false, "countsAsGeoSuccess": false }, "currentTop3DeliveryFeatures": { "capturedAt": "2026-08-12T22:10:27+00:00", "competitorObservationCount": 9, "distinctCompetitorUrlCount": 8, "successfulCompetitorFetchCount": 1, "failedFetchCount": 8, "targetFetchStatus": "success", "targetTableCount": 4, "targetJsonLdBlockCount": 8, "targetDownloadableLinkCount": 79, "targetObservedInUnbrandedResults": false, "jsonUrl": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-current-top3-delivery-features-2026-08-12.json", "csvUrl": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-current-top3-delivery-features-2026-08-12.csv", "targetReferenceCountsAsSearchVisibility": false, "provesRankingCause": false, "provesSourceQuality": false, "countsAsGeoSuccess": false } }, "bingSearchEvidence": { "engineId": "bing", "market": "CN", "language": "zh-CN", "accessMode": "anonymous", "observedAt": "2026-08-15T02:17:54Z", "baselineObservedAt": "2026-08-15T01:42:27Z", "releaseId": "20260815T020335Z-6f81ae955e76", "queryCount": 3, "organicPositionCount": 30, "distinctUrlCount": 28, "retainedUrlOccurrenceCount": 30, "enteredUrlOccurrenceCount": 0, "exitedUrlOccurrenceCount": 0, "rankChangedRetainedUrlOccurrenceCount": 4, "targetObservedQueryCount": 1, "brandDomainObservedQueryCount": 1, "queries": [ { "queryId": "rag-citation-provider-sources", "textZh": "千问 Kimi Gemini RAG 微调 引用来源", "currentResultCount": 10, "retainedUrlCount": 10, "enteredUrlCount": 0, "exitedUrlCount": 0, "rankChangedRetainedUrlCount": 0, "targetObserved": false, "brandDomainObserved": false }, { "queryId": "rag-citation-source-comparison", "textZh": "AI 搜索 RAG 微调 引用来源 对比", "currentResultCount": 10, "retainedUrlCount": 10, "enteredUrlCount": 0, "exitedUrlCount": 0, "rankChangedRetainedUrlCount": 4, "targetObserved": true, "brandDomainObserved": true }, { "queryId": "rag-citation-selection-sources", "textZh": "RAG 和微调选型 AI 引用哪些来源", "currentResultCount": 10, "retainedUrlCount": 10, "enteredUrlCount": 0, "exitedUrlCount": 0, "rankChangedRetainedUrlCount": 0, "targetObserved": false, "brandDomainObserved": false } ], "jsonUrl": "https://course.dragonai.tech/datasets/ai-search-rag-bing-current-query-delta-2026-08-15-r10.json", "csvUrl": "https://course.dragonai.tech/datasets/ai-search-rag-bing-current-query-delta-2026-08-15-r10.csv", "providerCallsMade": false, "countsAsPositiveSearchVisibility": true, "countsAsGeoSuccess": false }, "crossEngineSearchEvidence": { "engines": [ "baidu", "bing" ], "market": "CN", "queryCount": 3, "baiduPositionCount": 25, "bingPositionCount": 30, "baiduDistinctUrlCount": 22, "bingDistinctUrlCount": 28, "distinctUrlUnionCount": 49, "distinctUrlIntersectionCount": 1, "sameQuerySharedExactUrlRowCount": 1, "sameQuerySharedExactHostRowCount": 5, "sameQuerySharedExactHostCount": 2, "targetObservedQueryCount": 1, "brandObservedQueryCount": 1, "jsonUrl": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-bing-current-exact-url-overlap-2026-08-15-r2.json", "urlCsvUrl": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-bing-current-exact-url-overlap-2026-08-15-r2.csv", "hostCsvUrl": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-bing-current-exact-url-overlap-2026-08-15-r2-hosts.csv", "providerCallsMade": false, "countsAsPositiveSearchVisibility": true, "countsAsGeoSuccess": false }, "currentCrossSurfaceEvidence": { "definition": "Current strict Baidu organic URLs joined to current fixed-query AI citation URLs; hostname overlap is normalized separately from byte-identical exact URL overlap.", "baiduDistinctUrlCount": 22, "aiDistinctUrlCount": 50, "sharedHostnameCount": 3, "exactUrlOverlapCount": 2, "fullOuterJoinRowCount": 70, "targetObservedOnEitherSurface": false, "brandMentionedInAiRuns": false, "datasetUrl": "https://course.dragonai.tech/datasets/ai-search-rag-current-baidu-ai-overlap-2026-08-11.json", "hostnameCsvUrl": "https://course.dragonai.tech/datasets/ai-search-rag-current-baidu-ai-overlap-2026-08-11.csv", "urlFullOuterJoinCsvUrl": "https://course.dragonai.tech/datasets/ai-search-rag-current-baidu-ai-overlap-2026-08-11-urls.csv", "figureUrl": "https://course.dragonai.tech/datasets/ai-search-rag-current-baidu-ai-overlap-2026-08-11.svg", "countsAsPositiveSearchVisibility": false, "countsAsExternalLinkEvidence": false, "countsAsGeoSuccess": false }, "reproductionProtocol": { "phase": "postpublication", "providers": [ "qwen", "kimi", "gemini" ], "repeatPerProvider": 3, "comparisonBaselineReleaseId": "20260721T012841Z-5399a9c64530", "crawlReportReleaseBinding": "current_production_release", "requireBenchmarkReady": true, "providerCallsBeforeReady": false }, "lineage": { "immutableBaseAnswer": "https://course.dragonai.tech/research/ai-search-rag-finetuning-citations-2026-07/answer.json", "frozenTenObservationPredecessor": "https://course.dragonai.tech/datasets/ai-search-rag-ten-observation-answer-supplement-2026-07-31.json", "currentBaiduEvidence": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-comparable-observation-ledger-2026-08-14-r10.json", "currentBaiduUrlPersistence": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-serp-twenty-six-observation-persistence-2026-08-02.json", "currentBaiduQueryHostPersistence": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-query-host-persistence-twenty-six-observation-2026-08-02.json", "currentBaiduPersistentTop3SourceCards": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-persistent-top3-source-cards-twenty-six-observation-2026-08-02.json", "historicalBaiduBaseline": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-current-release-baseline-2026-08-01.json", "currentBaiduSnapshot": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-current-strict-snapshot-2026-08-11.json", "currentBaiduGeneratedAnswerEvidence": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-generated-answer-evidence-2026-07-30.json", "currentBaiduSelectionAnswerSourceLayers": "https://course.dragonai.tech/datasets/baidu-selection-answer-source-layer-gap-2026-08-13.json", "currentBaiduSelectionQueryDelta": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-selection-query-current-delta-2026-08-14-r5.json", "currentSelectionAuthorityAnswer": "https://course.dragonai.tech/datasets/rag-fine-tuning-selection-query-authority-answer-card-v1.json", "currentSelectionQueryBrowserObservation": "https://course.dragonai.tech/datasets/baidu-selection-query-browser-observation-2026-08-13t194800z.json", "currentCrossSurfaceEvidence": "https://course.dragonai.tech/datasets/ai-search-rag-current-baidu-ai-overlap-2026-08-11.json", "currentBaiduAdjacentResultChurn": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-current-release-adjacent-churn-2026-08-12.json", "currentTop3DeliveryFeatures": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-current-top3-delivery-features-2026-08-12.json", "currentBingQueryDelta": "https://course.dragonai.tech/datasets/ai-search-rag-bing-current-query-delta-2026-08-15-r10.json", "currentBaiduBingExactUrlOverlap": "https://course.dragonai.tech/datasets/ai-search-rag-baidu-bing-current-exact-url-overlap-2026-08-15-r2.json" }, "measurementBoundaries": { "providerCallsExecutedForThisUpdate": false, "countsAsCrawlEvidence": false, "countsAsIndexing": false, "countsAsPositiveSearchVisibility": false, "countsAsTargetRanking": false, "countsAsRankingCausality": false, "countsAsExternalLinkEvidence": false, "countsAsGeoSuccess": false, "rule": "This endpoint joins frozen provider observations with the latest comparable anonymous Baidu evidence. Zero target results and first-party publication do not prove indexing, ranking, citation eligibility, positive visibility, or GEO success." } }