Jia Li | Chlience Paper Archive

Jia Li

Author of FlashMemory-DeepSeek-V4 with paper-time affiliation HKUST(GZ). Homepage identifies Jia Li as an associate professor in Data Science and Analytics at HKUST(GZ), director of DSAIL, with PhD from CUHK and prior internships at Google AI and Tencent AI Lab.

1 archived notes X: not-found Homepage

Jul 05, 2026

FlashMemory-DeepSeek-V4: Lightning Index Ultra-Long Context via Lookahead Sparse Attention

FlashMemory DeepSeek V4 在 DeepSeek V4 Flash 的 HCA / CSA 结构上增加一个 lookahead Neural Memory Indexer，把历史 CSA chunk 放入 CPU cold pool，再按未来约 $64$ 个 token 的预测需求取回少量 chunk；它把平均物理 KV 占用降到 full context baseline 的 $13.5\%$，平均分数从 $76...

2606.09079-flashmemory-deepseek-v4-lookahead-sparse-attention KV Cache Sparse Attention Long Context

Related Notes

FlashMemory-DeepSeek-V4: Lightning Index Ultra-Long Context via Lookahead Sparse Attention