Unverified50% confidenceFactExact time
DeepSeek R1 achieves emergent long-range reasoning capabilities using only an ORM through the GRPO algorithm.
1
Sources
50%
Confidence
Long-term
Relevance
7/22/2026
First Seen
Sources
Related Claims
VerifiedDeepSeek-R1通过在代码和数学两个可验证域上大规模运行GRPO算法取得成功85% similarUnverifiedDeepSeek explored the GRPO algorithm as an attempt to improve upon the original RLHF framework.78% similarVerifiedGRPO由DeepSeek团队系统化提出并在训练DeepSeek-R1系列模型时得到大规模验证76% similarUnverifiedDeepSeek团队报告，仅用GRPO在数学和代码数据上进行强化学习而不经过监督微调，模型便自发出现反思、回溯验证等链式推理行为（aha moment）76% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/592332API
curl https://kongchang.com/api/v1/knowledge/claims/592332MCP
get_claim(id=592332)