概率感知,确定性权威:将模型生成的观测结果纳入经充分性检查的治理合约
Probabilistic Sensing, Deterministic Authority: Admitting Model-Produced Observations into Sufficiency-Checked Governance Contracts
浏览论文内容
中文总结 AI 辅助
该研究提出将模型感知的带分数观测结果纳入经充分性检查的治理合约,以联合界限定感知改变裁决的概率,在构造领域的测试中验证了该方法的有效性。
中文摘要 AI 辅助
当权威合约所需的领域仅存在于非结构化证据中时,模型可对其进行感知。我们仅将模型的输出作为带有分数的观测记录予以接纳。接纳策略基于保留集拟合,在声明的误报上限下,将每个分数映射为“是”“否”或“未知”,“未知”即拒绝。确定性、经充分性检查的合约据此作出决策。感知改变裁决结果的概率,被限定为合约所感知领域上的接纳错误率与未知率之和,这是由合约索引的联合界推理的实例化。最小化该估计界是在充分性合约间进行选择的有效成本模型。在一项针对两个构造领域、两个传感器族(36000次模型调用)的注册研究中,无任何单元反驳该界。本项目中首次出现了感知导致“拒绝转允许”的情况:21000个测试裁决中有13个,均来自3条矛盾记录;每个单元中经注册界限定的翻转均在该界之下。感知感知型选择在4项注册测试中均选了暴露度更低的合约。两个传感器的分数均具有信息性但未校准。正确性相对于声明的损失模型、候选表示及可达状态而言,所有领域均为构造的。
英文摘要
When a field that an authority contract needs exists only in unstructured evidence, a model can sense it. We admit the model's output only as an observation record with a score. An admission policy, with thresholds fitted on a held-out split at a declared false-positive ceiling, maps each score to true, false or unknown. Unknown denies. A deterministic, sufficiency-checked contract decides. The probability that sensing changes the verdict is bounded by the sum, over the contract's sensed fields, of the admitted-wrong and unknown rates. This is an instantiation of union-bound reasoning, indexed by the contract. Minimising the estimated bound is a valid cost model for choosing among sufficient contracts. In a registered study on two constructed domains with two sensor families (36,000 model calls), no cell refuted the bound. Deny-to-allow changes from sensing appeared for the first time in this programme: 13 of 21,000 test verdicts, all from 3 contradictory records; each flip in a cell with a registered bound lay under it. Sensing-aware selection picked the lower-exposure contract in 4 of 4 registered tests. Both sensors' scores were informative but not calibrated. Correctness is relative to the declared loss model, candidate representation and reachable states; all domains are constructed.
发表机构
- Universidad Torcuato Di Tella(托尔夸托·迪泰拉大学)
机构由 AI 辅助整理,请以论文原文为准。