The Forget-Retain Alignment Gap is introduced, a training-free predictor that scores an update's forget-retain alignment without running a relearning attack, and separates selective from dense updates more reliably than global distance, suggesting that weight selectivity better explains robustness than distance alone.
Yi Chen, Hanna Hsieh, Shu-Hong Liu et al.· 0 citations
This work places the hot set in HBM and the cold pool in HBF, forming a hot-cold KV hierarchy within the GPU memory tier, while flash-establishing HBF as a cold-tier complement to HBM rather than its replacement.
Jongjin Baek, Won Ji, Seungjae Yoo et al.· IEEE computer architecture l...· 0 citations
APT, a software-hardware co-designed accelerator for high-resolution DiTs that leverages attention probabilities as a unified importance metric to jointly optimize computation through fine-grained pruning and adaptive precision scaling, and is evaluated on SOTA DiT models including PixArt, Stable Diffusion 3, and FLUX.
Sungyeob Yoo, Seeyeon Kim, Joonyong Park et al.· 0 citations
This work explores an HBF organization that simultaneously exploits two independent expert-delivery routes: a direct path that transfers expert weights from HBF to the GPU and a relay path that transfers them from HBF through the HBM base die to the GPU.
Seeyeon Kim, Juhyeong Jin, Joo-Young Kim· 4 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
