4 papers
Residual-Mass Accounting for Partial-KV Decoding
Yasuto Hoshi, Daisuke Miyashita, Jun Deguchi
We study a controlled partial-KV decoding setting in which exact unnormalized softmax contributions are computed for sink/tail anchors and a retrieved token set, while the remainin…
Revolutionizing Long-Term Memory in AI: New Horizons with High-Capacity and High-Speed Storage
Hiroaki Yamanaka, Daisuke Miyashita, Takashi Toi +3
Driven by our mission of "uplifting the world with memory," this paper explores the design concept of "memory" that is essential for achieving artificial superintelligence (ASI). R…
AiSAQ: All-in-Storage ANNS with Product Quantization for DRAM-free Information Retrieval
Kento Tatsuno, Daisuke Miyashita, Taiga Ikeda +3
Graph-based approximate nearest neighbor search (ANNS) algorithms work effectively against large-scale vector retrieval. Among such methods, DiskANN achieves good recall-speed trad…
On Storage Neural Network Augmented Approximate Nearest Neighbor Search
Taiga Ikeda, Daisuke Miyashita, Jun Deguchi
Large-scale approximate nearest neighbor search (ANN) has been gaining attention along with the latest machine learning researches employing ANNs. If the data is too large to fit i…