Unverified50% confidenceFactExact time
OpenAI、Google DeepMind等机构的后续研究也证实了谄媚倾向是当前训练范式下的普遍特征
1
Sources
50%
Confidence
Long-term
Relevance
8/26/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedOpenAI在InstructGPT论文中证明，少量高质量人类偏好数据对模型对齐的贡献远超大规模无监督训练74% similarUnverifiedResearch from OpenAI, Anthropic, and other institutions indirectly confirms that alignment training significantly compresses a model's performance space in creative writing, complex reasoning, and open-ended dialogue.71% similarUnverifiedResearch by Google DeepMind and other institutions in 2023 confirmed that a model's memorization behavior is closely related to the number of repetitions of training data.69% similarUnverified行为克隆预训练解决了强化学习的冷启动问题：纯强化学习在稀疏奖励环境中往往需要数亿次交互才能学到基本可用策略67% similarUnverified如果模型能理解行为准则背后的深层原因，即使面对训练中从未遇到的全新场景也能做出合理的行为选择，实现可泛化的对齐66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/805075API
curl https://kongchang.com/api/v1/knowledge/claims/805075MCP
get_claim(id=805075)