Unverified50% confidenceFactExact time
ContractNLI is a dataset built by Stanford for NLI tasks that contains real contradictions; its task format is hypothesis-to-clause rather than the symmetric clause-to-clause comparison.
1
Sources
50%
Confidence
Long-term
Relevance
9/12/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedContract Testing 理念（如 Pact 框架）可用于 AI Skill 自动维护服务间契约文件，降低微服务集成测试协调成本61% similarUnverifiedRaft 共识算法由 Stanford 于 2014 年发表，以可理解性为设计目标，将共识过程分解为 Leader 选举、日志复制和安全性保证三个模块54% similarUnverifiedCodex在自动化缺陷修复场景中的工作流包括：静态分析触发缺陷报告→AI代理读取代码上下文→生成候选修复方案→沙箱中运行测试验证→提交Pull Request供人工审核54% similarUnverifiedOpenAI的合同审查智能体批量读取协议并结构化数据，识别非标准条款以确保符合GAAP、ASC 606等收入确认规则52% similarUnverifiedHumanEval、MBPP等传统AI基准本质是函数级黑盒测试，无法覆盖多模块协作、状态持久化和构建流程等工程维度52% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/910543API
curl https://kongchang.com/api/v1/knowledge/claims/910543MCP
get_claim(id=910543)