一作ICLR 2026Poster· Deep Learning / Attention Mechanisms
LycheeDecode: Accelerating Long-Context LLM Inference via Hybrid-Head Sparse Decoding
Gang Lin · dongfang li · Zhuoen Chen · Yukun Shi · Xuhui Chen · Baotian Hu · Min Zhang
首次获取该候选人的完整档案及 StartLux 评估共 20 积分；每位用户单独解锁，之后查看免费。主动刷新 20 积分。
实时查找电子邮箱，费用不超过20积分；未找到邮箱不扣费。
Gang Lin · dongfang li · Zhuoen Chen · Yukun Shi · Xuhui Chen · Baotian Hu · Min Zhang
dongfang li · Zixuan Liu · Gang Lin · Baotian Hu · Min Zhang
The quadratic complexity of the attention mechanism and the substantial memory footprint of the Key-Value (KV) cache present severe computational and memory challenges for Large Language Models (LLMs) processing long contexts. Existing retrieval-based methods often compromise semantic integrity t...