Expired50% confidenceFactExact time
DeepSeek V4 adopted the Muon optimizer improved by the Kimi (Moonshot AI) team, which is a second-order optimization algorithm based on Nesterov momentum.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/22/2026
First Seen
This claim has expired. It is kept for the record and no longer describes the current state.
Valid until: 10/20/2026(expired)
Sources
Related Claims
UnverifiedDeepSeek V4采用MUON优化器替代AdamW进行训练73% similarUnverifiedDeepSeek-V2 and V3 adopt a MoE architecture similar to Kimi K2.7, and introduce optimizations such as shared experts and fine-grained expert splitting.71% similarUnverifiedDeepSeek V4系列采用了MLA、MoE路由优化以及NSA等创新架构技术71% similarUnverifiedThe core innovations of DeepSeek V4 include continued optimization of the MoE (Mixture of Experts) architecture and deep engineering tuning of the training pipeline.71% similarVerifiedDeepSeek V4 adopts a Mixture of Experts (MoE) architecture.71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/590222API
curl https://kongchang.com/api/v1/knowledge/claims/590222MCP
get_claim(id=590222)