Unverified50% confidenceOpinionExact time
Critics point out that the low-probability, high-impact framework has room for manipulation—probabilities can be arbitrarily lowered and impacts can be arbitrarily raised—making it easy to fall into Pascal's-wager-style argumentation traps.
1
Sources
50%
Confidence
Long-term
Relevance
9/17/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedDenser relevance judgments reduce the probability that correct results retrieved by a model are mistakenly judged as irrelevant, thereby improving evaluation reliability.74% similarUnverified负向约束通过明确排除不期望的输出收窄概率空间，与正向指令同等重要，可减少模型创造性发挥的偏差73% similarUnverifiedOver-refusal strategies reduce risk but harm model utility, especially in scenarios such as professional development and security research that require touching sensitive boundaries.73% similarUnverifiedThe low-probability, high-impact risk framework originates from expected utility theory and catastrophic risk research.73% similarUnverified越接近高风险动作，越应该减少Mock、增加真实交互，测试应分层展开为单元层、集成层、端到端层72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/929325API
curl https://kongchang.com/api/v1/knowledge/claims/929325MCP
get_claim(id=929325)