Unverified70% confidenceOpinionTime unknown
大模型对指令的敏感度呈非线性分布——主干指令的微小改动可能引发全局行为漂移，而在末尾追加具体的边界条件则相对安全
1
Sources
70%
Confidence
Long-term
Relevance
6/1/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedResearchers have found that fine-tuning only a small number of model parameters is sufficient to remove safety alignment (de-alignment attacks), and that certain jailbreak prompts can transfer across different models.78% similarUnverified底层模型更换后，即便微小的输出分布变化也可能导致此前校准的安全护栏过度拦截或防护失效，需要重新标注和调试77% 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76% similarUnverified通过'不要按常规套路'这样的约束性提示词，本质上是在调整模型的采样分布，降低高频方案的输出权重，迫使模型探索低概率但可能更具创新性的输出区域75% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/24808API
curl https://kongchang.com/api/v1/knowledge/claims/24808MCP
get_claim(id=24808)