Outeffhop: A principled outlier-efficient attention layer from dense associative memory models

H Luo, JYC Hu, PH Chang, HY Chen, W Li… - Workshop on Efficient …, 2024 - openreview.net
We introduce a principled approach to Outlier-Efficient Attention Layers via associative
memory models to reduce outlier emergence in large transformer-based model. Our main …

Fast adaptation and robust quantization of multi-modal foundation models from associative memory: A case study in speechLM

S Wu, YJ Lu, H Luo, JYC Hu, J Wang… - Workshop on Efficient …, 2024 - openreview.net
We present a preliminary investigation into the outlier problem within the multi-modal
foundation model with a focus on SpeechLM. Specifically, we consider SpeechLM models …

Privacy-Preserving Analytics of Human Mobility Data

A Kapp - 2024 - tud.qucosa.de
Abstract (EN) In recent years, human mobility data has been increasingly collected and
stored, mainly due to the ubiquitous use of smartphones, producing a constant stream of …