How models fail in deployment
Distribution shift, spurious correlations and feedback loops are explained with public case records, so oversight bodies know what to ask about a system’s behaviour after deployment.
GRF · 人工智能与数字治理
Artificial intelligence and machine learning are the 系统 this field governs. On this platform the methods are explained so that oversight bodies, legislators and the public understand what models do, how they fail and what evidence about them can be trusted.
Artificial intelligence and machine learning develop algorithms that learn patterns from data to classify, predict or generate outputs. They are increasingly used for hazard detection, forecasting and decision support, and are themselves a source of risk. In artificial intelligence, digital public infrastructure and cyber and digital resilience, machine learning can scale detection and forecasting across interconnected 系统. Its own failure modes must be weighed when automated outputs inform consequential decisions. The canonical description is on STEM Commons.
Distribution shift, spurious correlations and feedback loops are explained with public case records, so oversight bodies know what to ask about a system’s behaviour after deployment.
Model cards, data sheets and system cards are read for what they disclose and omit about training data, evaluation and limits.
How agents chain model calls, tools and memory is described, so that permission and liability questions can be placed at the right point.
团队遵循该方法时会使用的界面:其指标、状态与决策。名称为示例,数值仅作示意。
AI and digital governance · Artificial intelligence · Run R-6213
| 全部 | 0.85 | 0.04 | 4407 | 通过 |
| 区域 A | 0.78 | 0.04 | 1829 | 审核 |
| 区域 B | 0.65 | 0.07 | 444 | 审核 |
| 稀有类别 | 0.74 | 0.14 | 124 | 低于标准 |
| 数据 | 角色 | 标签 |
|---|---|---|
| Model and system cards | Disclosed capabilities and limits | 已声明 |
| Benchmark results | Comparative capability | 已报告 |
| Incident records | Observed failures | 已报告 |
| Research literature | 方法与证据 | 已报告 |
三类工具在你设定的覆盖率和遵从率下与基线并列比较,每类都有自己的滞后和采用速度;效应大小是教学用数值,应以证据替换。情景:关于未来的具名假设,不是预测。
控制措施
读数
| 年份 | 基线 | 监管 | 经济激励 | 信息与能力 |
|---|---|---|---|---|
| 0 | 100.0 | 99.5 | 98.8 | 99.1 |
| 1 | 102.0 | 101.2 | 100.1 | 99.8 |
| 2 | 104.0 | 102.5 | 101.1 | 99.9 |
| 3 | 106.1 | 103.5 | 101.6 | 100.0 |
| 4 | 108.2 | 103.7 | 101.8 | 100.7 |
| 5 | 110.4 | 103.2 | 101.7 | 102.1 |
| 6 | 112.6 | 102.0 | 101.6 | 103.9 |
| 7 | 114.9 | 100.4 | 101.6 | 105.9 |
| 8 | 117.2 | 99.2 | 102.1 | 108.0 |
| 9 | 119.5 | 98.6 | 102.9 | 110.1 |
| 10 | 121.9 | 98.9 | 104.2 | 112.3 |
| 11 | 124.3 | 99.7 | 105.7 | 114.5 |
| 12 | 126.8 | 101.1 | 107.4 | 116.8 |
| 13 | 129.4 | 102.7 | 109.4 | 119.2 |
| 14 | 131.9 | 104.5 | 111.4 | 121.6 |
| 15 | 134.6 | 106.5 | 113.5 | 124.0 |
| 16 | 137.3 | 108.5 | 115.7 | 126.5 |
| 17 | 140.0 | 110.7 | 118.0 | 129.0 |
| 18 | 142.8 | 112.9 | 120.4 | 131.6 |
| 19 | 145.7 | 115.1 | 122.8 | 134.2 |
| 20 | 148.6 | 117.4 | 125.2 | 136.9 |
The model runs on the teaching values set by the controls. To calibrate it for a place, read it against these admitted layers and series, which this page’s map can draw:
贯穿六大支柱,由 Nexus Consortium 牵头
模型在尊重空间、时间与群体结构的留出数据上进行评估。在测试准确性的同时,也测试稳健性、校准与公平性。
| 核查 | 如何开展 |
|---|---|
| Claims against evidence | Capability claims are checked against published evaluations |
| Context of benchmarks | Benchmark results are read with their task and data limits stated |
| Deployment conditions | Behaviour is described for the conditions in which the system is used |
当部署条件与训练数据不同时,性能会下降。模型可能固化偏见,并在罕见事件上毫无明显预警地失效。
一项文书从授权到评估所经历的步骤,每一步都标明负责运行的支柱,以及仍由当局作出的决定。这是带有明示说明性数值的教学模型,不是预测。
贯穿六大支柱,由 Nexus Consortium 牵头
政策制定者、外交官或发展机构所使用的文书、授权、行为方和问责安排,并注明设定它们的协定。结构遵循所列框架。所绘制的任何状态或进展均为说明性,并非记录。
The kinds of instrument available on Artificial intelligence and machine learning for AI and digital governance, by level of government and cooperation.
| 工具类型 | 地方 | 国家 | 区域 | 全球 |
|---|---|---|---|---|
| 监管型 | 分区、建筑规范 | 法律、标准、许可 | 统一规则 | 公约与条约 |
| 经济 | 地方税费、拨款 | 预算项目、税收、保险计划 | 风险池,区域基金 | 融资机制 |
| 信息 | 地方预警 | 披露、预警、统计 | 数据共享安排 | 全球监测与报告 |
| 制度型 | 地方委员会 | 牵头机构,协调理事会 | 区域机构与秘书处 | 多边授权 |
| 合作型 | 社区协议 | 伙伴关系契约 | 谅解备忘录与联合计划 | 协定与共同承诺 |
报告 is the Consortium’s quarterly publication. What you know about this subject can become evidence, a case or a correction in the next edition, credited to you and published under CC BY 4.0 where it is your own work.
切勿发送个人数据、敏感基础设施细节或您无权分享的任何内容。贡献在使用前会经过审查;发送贡献不赋予署名权或编辑权,赞助也绝不能换取研究结论。
本网站上的地图、数据和图表仅作为背景和一般信息展示。它们不是建议、预测或预警;信息的局限见免责声明和数据说明。
This site uses only strictly necessary cookies to work and to remember your privacy choices. Nothing loads from other companies, and we never sell or share your personal information.
This site uses only strictly necessary cookies to work and to remember your privacy choices. Nothing loads from other companies, and we never sell or share your personal information. We run no advertising. You can review your choices at any time through Privacy choices at the foot of every page.