ConsensusConsensus
Consensus 回答「科学界怎么说」：Consensus Meter 把多篇研究的结论极性统计成摘要—— 快速判断实证领域共识方向；但计数不等权，Meter 不是定论。
Consensus answers "what does the scientific literature say": the Consensus Meter tallies conclusion polarity across studies into a summary — a fast read on empirical consensus, but the count is unweighted and the Meter is not a verdict.
适合与不适合When to use it, when not to
适合：实证领域共识速查、文献初筛、结论极性判断。 不适合：深度综述（抽取轻于 Elicit）、中文文献、把 Meter 当答案终点。
"What does science say" questions (medical/nutrition/climate and other empirical fields), fast conclusion-polarity reads, literature first-pass screening. Not for: deep systematic reviews (the extraction layer is lighter than Elicit), Chinese literature as the main axis, treating the Meter as settled truth.
中文生态China availability 国产 / 国内可直接用 / 备案状态 · 快照domestic / direct access / CAC filing · snapshot 2026-10-22
- 出品属地Origin
- 国外出品foreign-made Consensus（美国）。Consensus (US).
- 国内使用Direct access
- 大陆可直连direct OK consensus.app 大陆直连实测 200（2026-10-22）。consensus.app measured 200 direct from mainland (2026-10-22).
- 备案状态CAC filing
- 未见于备案名单not in filing list：未见于网信办公示名单（2026-10-22 检索）；该产品未面向大陆公众提供服务，备案不适用。not in the CAC public registry (searched 2026-10-22); the product does not serve the mainland public, so filing does not apply.
备案信息以网信办《生成式人工智能服务已备案信息》公示名单为准（按批次滚动更新，本站只抄编号不解读）；「大陆可直连」是本站从大陆网络直连实测的结果，不是厂商承诺。GitHub 类站点当日实测有波动，口径写在依据里。Filing status is quoted from the CAC public registry (updated batch by batch; this site copies the number and does not interpret it). "Direct access" is this site's own reachability measurement from a mainland network, not a vendor promise. GitHub-hosted sites fluctuated on the measurement day — the note says so.
固定坐标系Fixed coordinate system
8 个维度dimensions, 与同分区其他对象逐项可比each comparable item by item against others in the same section.
- 模型与开放条件Model & openness
科学文献检索引擎+结论抽取：自建论文索引（2 亿+ 篇叙事，官方站点口径——未逐条核验） +Consensus Meter（多篇研究的结论极性统计，产品招牌）。
Scientific-literature retrieval engine + conclusion extraction: a self-built paper index (200M+ papers narrative, official-site framing — not verified item by item) + Consensus Meter (polarity statistics across multiple studies' conclusions, the product's signature).
- 运行位置Where it runs
网页（consensus.app）为主入口；文献检索与结论生成都在厂商云端完成，未见官方声明的本地形态。
Web (consensus.app) as the main entry; literature retrieval and conclusion generation both happen on the vendor's cloud; no officially stated local form seen.
- 本地文件Local files
搜索历史与高亮标注存账户侧云端；PDF/引用导出归本地。未见官方声明的其他本机数据面。
Search history and highlight annotations are stored on the account-side cloud; PDF/citation exports count as local. No other local data surfaces officially stated seen.
- 关机后的任务Tasks after shutdown
检索与 Meter 统计即时返回；未见官方任务队列或「关机续跑」类后台任务面（未逐条核验）。
Retrieval and Meter statistics return instantly; no official task queue or "keep-running-after-shutdown" style background task surface seen (not verified item by item).
- 工具与扩展Tools & extensions
Consensus Meter 是招牌：对「科学界对 X 的共识方向」给统计式摘要（支持/反对/混合计数）； 按研究类型/样本量筛选；AI 综合答案带引用；PDF 高亮问答（Pro 线，官方产品面）。
Consensus Meter is the signature: a statistical summary of "which way science leans on X" (support/oppose/mixed counts); filtering by study type/sample size; AI synthesized answers with citations; PDF highlight Q&A (Pro line, official product surface).
- 上下文与记忆Context & memory
以搜索历史与文件夹组织研究项目；跨会话记忆与用户画像未核验到官方声明（未逐条核验）。
Research projects organized via search history and folders; cross-session memory and user profiling not verified to an official statement (not verified item by item).
- 权限与限制Permissions & limits
账号+订阅体系（D9：价格不采）；免费档功能限制（未逐条核验）；训练用途开关未逐条核验。
Account + subscription system (D9: pricing not collected); free-tier feature limits not verified item by item; the training-use switch not verified item by item.
- 适合什么任务Best for
适合：「科学界怎么说」类问题（医学/营养/气候等实证领域）、快速结论极性判断、 文献初筛。 不适合：深度系统性综述（抽取面比 Elicit 轻）、中文文献主打、把 Meter 当定论。
Fits: "what does science say" questions (medicine/nutrition/climate and other empirical fields), quick polarity calls, literature triage. Does not fit: deep systematic reviews (its extraction surface is lighter than Elicit's), Chinese-literature-first, treating the Meter as settled science.
头号误解Top misconception
- Consensus Meter 是「论文计数的极性统计」不是「科学定论」——论文质量不等权，读 Meter 要看底层论文The Consensus Meter is an unweighted polarity count over papers, not a scientific verdict — paper quality is not equal-weighted; read the Meter together with the underlying papers
- 检索覆盖偏英文实证文献——人文/中文/非期刊来源覆盖有限，如实按领域评估Coverage skews to English empirical literature — humanities, Chinese, and non-journal sources are limited; evaluate field by field
- consensus.app 对 curl 返回 CF 403——本站核验方法受限如实标注，不是站点问题consensus.app returns a Cloudflare 403 to curl — our verification method is limited and noted honestly; it is not a site problem
未知项清单Known unknowns
- 索引规模与来源构成（官方叙事未逐条核验）
- ToS 细节（CF 403 未能直抓）
- 训练用途开关（未逐条核验）
- 中文文献覆盖（未核验到官方声明）
- Index size and source composition (official narrative not verified item by item)
- ToS details (CF 403 blocked direct fetch)
- The training-use switch (not verified item by item)
- Chinese-literature coverage (no official statement verified)
证据来源Evidence sources
判断可回到以下一手源复核。本站核验日Every judgement can be rechecked against the primary sources below. Verified on 2026-10-22，内容更新日content updated on 2026-10-22。
| 类型Type | 名称Name | 链接Link |
|---|---|---|
| official | Consensus · 官网（产品与能力叙事）Consensus · official site (product and capability narrative) | https://consensus.app |
| official | Consensus · Terms（官方条款，直抓 CF 403，2026-10-22）Consensus · Terms (official terms; direct fetch hit CF 403, 2026-10-22) | https://consensus.app/terms/ |
| changelog | Consensus · Blog（官方更新线，changelog；直抓 CF 403）Consensus · Blog (official update line, changelog; direct fetch hit CF 403) | https://consensus.app/blog/ |
实测记录Measured runs
本档未跑检索质量评测；结论只覆盖产品能力与来源披露。
实测建议（三条，都指向官方可读页没给的东西）：
- Meter 对争议话题的极性统计与人工综述的偏差
- 按样本量/研究类型筛选后的统计变化
- 免费档与 Pro 的功能边界
This site has not completed measured runs. The coding-oriented test protocol is at protocols/coding/_protocol.md (the coding task suite does not apply to creation products).
Suggested measurements (three, all pointing at things the official readable pages do not give):
- How the Meter's polarity stats on contested topics deviate from human reviews
- How the statistics shift after filtering by sample size/study type
- The feature boundary between the free tier and Pro
相关条目Related entries
- ElicitElicit同为科学文献向：Consensus 速查极性，Elicit 深度抽取综述。Also scientific-literature oriented: Consensus gives quick polarity reads, Elicit does deep extraction for reviews.
- NotebookLMNotebookLM互补：Consensus 找文献给极性，NotebookLM 在选定文献内深挖。Complementary: Consensus finds literature and gives polarity, NotebookLM digs deep inside the sources you pick.
- 秘塔 AI 搜索Metaso AI Search中文学术对照：秘塔学术档检索中文库，Consensus 英文实证池。Chinese academic counterpart: Metaso's academic tier searches Chinese libraries, Consensus works the English empirical pool.
- ChatGPTChatGPT另一层 · AI 对话助手Other layer · Chat assistants
本页由This page is generated from the ai-handbook 数据层生成（CC BY 4.0）。data layer (CC BY 4.0).
方法论与坐标系定义见仓库内For the methodology and coordinate definitions, see METHODOLOGY.md.