GitHub

Triton GDN kernel produces garbled text (foreign language token mixing) for dense Qwen 3.5 models

The Triton GDN (Gated DeltaNet) kernel produces garbled text with foreign language tokens intermixed when running dense Qwen 3.5 models (e.g., 9B). The same model produces clean output with both ...
IEEE

SeMv-3D: Toward Concurrency of Semantic and Multi-View Consistency in General Text-to-3D Generation

Abstract: General Text-to-3D (GT23D) generation is crucial for creating diverse 3D content across objects and scenes, yet it faces two key challenges: 1) ensuring semantic consistency between input ...
GitHub

lightx2v_kernel_xpu

onednn_w8a16_fp8(x, qweight, scales[, bias]) W8A16 GEMM — fp16/bf16 activations × FP8_E4M3 weights, per-column scale onednn_w4a16(x, weight, scales, zeros[, bias]) W4A16 GEMM — fp16/bf16 activations × ...
IEEE

BiSeR-LMA: A Bidirectional Semantic Reasoning and Large Model Enhancement Approach for Text-Video Cross-Modal Retrieval

Abstract: Video, as an information carrier, provides a vast amount of important information to people. Therefore, the method of obtaining video becomes particularly important, which drives the ...