[KongchangAI]
Unverified50% confidenceOpinionExact time

不同模型在抗诱导能力上的差异很大程度上源于后训练阶段(post-training)的策略差异

1
Sources
50%
Confidence
Long-term
Relevance
8/19/2026
First Seen

Sources

Related Claims

Cite This Claim

Stable URI
https://kongchang.com/claim/774276
API
curl https://kongchang.com/api/v1/knowledge/claims/774276
MCP
get_claim(id=774276)