[CPU] Add GPT-OSS model optimization for CPU (#16775)

Co-authored-by: mingfeima <mingfei.ma@intel.com>
Co-authored-by: jianan-gu <jianan.gu@intel.com>

sgl-kernel/csrc/cpu/qkv_proj.cpp

@@ -210,6 +210,7 @@ void segment_gemm_kernel_impl(
           /*   C */ C + mb_start * ldc + local_nb_start,
           /* Btmp*/ Btmp + tid * BLOCK_N * K,
           /* Ctmp*/ Ctmp,
+          /*Bbias*/ nullptr,
           /*  Bs */ Bs + (new_nb / blocks_n_per_group) * scale_size_K,
           /*   M */ mb_size,
           /*   N */ nb_size,