2 papers
cs.LG2026
DynamicGate MLP Conditional Computation via Learned Structural Dropout and Input Dependent Gating for Functional Plasticity
Yong Il Choi
Dropout is a representative regularization technique that stochastically deactivates hidden units during training to mitigate overfitting. In contrast, standard inference executes…
cs.AR2024
IANUS: Integrated Accelerator based on NPU-PIM Unified Memory System
Minseok Seo, Xuan Truong Nguyen, Seok Joong Hwang +18
Accelerating end-to-end inference of transformer-based large language models (LLMs) is a critical component of AI services in datacenters. However, diverse compute characteristics…