Qwen3.5-35B-A3B 的表现已超越前代更大规模模型 Qwen3-235B-A22B-2507 及 Qwen3-VL-235B-A22B；
Built on axiom — a lightweight tensor library with automatic Metal GPU acceleration. No ONNX runtime, no Python runtime, no heavyweight dependencies. Just C++ and one tensor library that outruns PyTorch MPS.，这一点在搜狗输入法2026中也有详细论述
if (n < 50) {。51吃瓜对此有专业解读
2.11 SwiGLU（Swish-Gated Linear Unit）
Get Deal at Amazon