Unverified50% confidenceSolutionExact time
当黄金标准关系标签在候选列表中总是排在第一位时，模型会学会投机取巧的规则即选第一个候选，而非学习真正的语义关系
1
Sources
50%
Confidence
Long-term
Relevance
8/5/2026
First Seen
Sources
Related Claims
Unverified选型应按任务结构而非模型名气选择，若三问（多步执行且可能改计划、错误成本高、结果可验证）都为是则优先用Opus 5，只有一个为是则先用更便宜的模型62% similarUnverifiedChoosing a model cannot rely solely on benchmark scores; the model's willingness to execute real-world tasks is the key practical value.62% similarUnverifiedAssigning each agent a dedicated capability boundary and prompt often outperforms having a single model handle everything, a principle that is connected to the effectiveness of chain-of-thought prompting.61% similarVerified跑分榜单衡量的是模型考试能力而非解决真实问题的能力61% similarUnverified模型评测正在从传统知识问答转向Agent能力与实际编码能力的比拼60% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/692517API
curl https://kongchang.com/api/v1/knowledge/claims/692517MCP
get_claim(id=692517)