AI 导读
仅 6B 激活参数,Qwen3.8-Flash-Next-Base 在 14 项基准中登顶 8 项,包括 MMLU-Pro、SuperGPQA、BBH 和 GSM8K。其余基准上,它与 Qwen3.7-Plus-Base 相比仍具竞争力。其 51B N-gram 嵌入参数采用确定性查找,不增加每 token 的矩阵乘法开销。
正文
With only 6B active parameters, Qwen3.8-Flash-Next-Base tops 8 of 14 benchmarks, including MMLU-Pro, SuperGPQA, BBH and GSM8K. And it remains competitive with Qwen3.7-Plus-Base on the rest. Its 51B N-gram embedding parameters use deterministic lookups, adding no per-token matrix-multiplication cost.
来源:@Alibaba_Qwen · x.com