Unverified85% confidenceFactTime unknown
RLHF对齐存在'过度对齐'问题,模型在完全无害的场景下也会过度谨慎,频繁拒绝正常的创意写作和角色扮演请求
1
Sources
85%
Confidence
Long-term
Relevance
6/3/2026
First Seen
Sources
高自由度AI伴侣聊天软件实测:无限制角色扮演对话体验
bilibili风铃草不夜侯
Related Entities
Cite This Claim
Stable URI
https://kongchang.com/claim/38474API
curl https://kongchang.com/api/v1/knowledge/claims/38474MCP
get_claim(id=38474)