Showing cs.LGShow all
2 papers · 1 filter
cs.LG2026
Don't be so Stief! Learning KV Cache low-rank approximation over the Stiefel manifold
Luca Benfenati, Matteo Risso, Andrea Vannozzi +5
Key-value (KV) caching enables fast autoregressive decoding but at long contexts becomes a dominant bottleneck in High Bandwidth Memory (HBM) capacity and bandwidth. A common mitig…
cs.LG2026
TyphoonMLA: A Mixed Naive-Absorb MLA Kernel For Shared Prefix
Ahmet Caner Yüzügüler, Ahmet Ãelik, Jiawei Zhuang +1
Multi-Head Latent Attention (MLA) is a recent attention mechanism adopted in state-of-the-art LLMs such as DeepSeek-v3 and Kimi K2. Thanks to its novel formulation, MLA allows two…