Primary sourceOpenAI
View sourceEvent ·Single source
OpenAI与ApolloAI发布奖励追求新研究及测量方法Contrastive SDF
OpenAI与ApolloAI合作发布新研究，探讨奖励追求行为（模型倾向于遵循评分者奖励而非用户意图）。提出新方法Contrastive SDF，量化模型信念对行为的影响程度。研究在alignment.openai.com上公开，包含具体实验和基准测试。
Current read
OpenAI联手ApolloAI，用Contrastive SDF方法测量模型是否在讨好评分者而非用户，搞对齐的必看。
2 sources76° AI heatLast updated 2026/7/21 18:05:38
Evidence
2 sourcesRelated source 1arXiv: OpenAI
View sourceConflict check
Dedup data only shows that these sources discuss the same event, not that they agree. There is no structured supporting or refuting evidence yet, so no conflict is inferred automatically.