minimax-m3.1IMAGE
质疑 M3.1 Flash 只是 M3 改名:稀疏化注意力加量化
用户怀疑 M3.1 Flash 实为 M3 改名,只多稀疏化 3 层注意力、量化并叠加 DSpark,认为从营销角度说得通但仍感意外。
Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)
@teortaxesTex5 天前收录原推发布于 2026.10.23


Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞) @teortaxesTex原推发布于 2026.10.23
Bro they actually rebranded M3 as M3.1 Flash, just sparsified the attn in 3 more layers, quantized it and added DSpark on top I guess it's fair in a marketing sense, M3 is on the scale of today's Flash models, but… I didn't believe this
这条是引用着另一条写的:「MiniMax M3.1 Flash 开源权重:428B 总参、23B 激活」
x.com/teortaxestex/status/2113447334021644595采集于 2026.10.23 12:09