Authors

Luotian Huang

Publications 2

Preprint Open access

WavePrune: One period is often enough for RoPE

Guancheng Du, Luotian Huang, Shaowen Wang et al. · 2026

Rotary Position Embedding (RoPE) encodes token positions by rotating each two-dimensional channel of the query and key vectors at a channel-specific frequency, making the attention logits invariant to a common shift of positions. However, this rotation is periodic, and it leads to position aliasing where relative posit …

Preprint Open access

DeepSeek-V4.1-Flash: Pushing the Limits of KV Cache Compression

DeepSeek-AI, Anyi Xu, B. Li et al. · 2026

The widespread adoption of long-horizon agents has made model workloads increasingly input-heavy. Although prior work has substantially reduced the cost of long-context computation, prefill remains computationally expensive, and large KV caches continue to strain HBM and SSD capacity and data-transfer bandwidth. Togeth …

Co-authors

Guancheng Du Shaowen Wang Si Li Kaifeng Lyu DeepSeek-AI Anyi Xu B. Li Bangcai Lin Bing Xue BingCheng Xian Bingzheng Xu Bochao Wu