Unverified50% confidenceFactExact time
Sage-A1.5 uses a MoE architecture, with a total parameter count of 35B and about 3B parameters activated per inference.
1
Sources
50%
Confidence
Long-term
Relevance
10/29/2026
First Seen
Sources
Related Entities
Related Claims
Unverified千问3.6的35B MOE模型每次只激活其中3B的参数80% similarVerifiedOrnith 1.5 35B-A3B uses a MoE architecture with a total parameter count of approximately 35B, activating only approximately 3B parameters per inference.80% similarUnverifiedPlayer-1-35B-A3B adopts an MoE architecture, with a total parameter count of 35B but only about 3B parameters activated per token.77% similarUnverifiedMoE模型中'35B'指总参数量，每个token只激活少量Expert（通常2-4个），实际计算量可能仅相当于6-7B的Dense模型74% similarUnverifiedThe MoE architecture activates only a small fraction of the 2.4T parameters during inference (typically 10–20%).74% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1110188API
curl https://kongchang.com/api/v1/knowledge/claims/1110188MCP
get_claim(id=1110188)