Home /
Research
Showing 163 - 168 / 904
Graph Random Neural Networks for Semi-Supervised Learning on Graphs
WenzhengFeng, JieZhang, YuxiaoDong....
Published date-12/01/2020
DataAugmentation, NodeClassification
We study the problem of semi-supervised learning on graphs, for which graph neural networks (GNNs) have been extensively explored. However, most existing GNNs inherently suffer from the limitations of over-smoothing, …
Practical Low-Rank Communication Compression in Decentralized Deep Learning
ThijsVogels, SaiPraneethKarimireddy, MartinJaggi....
Published date-12/01/2020
Lossy gradient compression has become a practical tool to overcome the communication bottleneck in centrally coordinated distributed training of machine learning models. However, algorithms for decentralized training with compressed communication …
Rel3D: A Minimally Contrastive Benchmark for Grounding Spatial Relations in 3D
AnkitGoyal, KaiyuYang, DaweiYang....
Published date-12/01/2020
Understanding spatial relations (e.g., laptop on table) in visual input is important for both humans and robots. Existing datasets are insufficient as they lack large-scale, high-quality 3D ground truth information, …
H-Mem: Harnessing synaptic plasticity with Hebbian Memory Networks
ThomasLimbacher, RobertLegenstein....
Published date-12/01/2020
QuestionAnswering
The ability to base current computations on memories from the past is critical for many cognitive tasks such as story understanding. Hebbian-type synaptic plasticity is believed to underlie the retention …
Learning to summarize with human feedback
NisanStiennon, LongOuyang, JeffreyWu....
Published date-12/01/2020
As language models become more powerful, training and evaluation are increasingly bottlenecked by the data and metrics used for a particular task. For example, summarization models are often trained to …
Stochastic Deep Gaussian Processes over Graphs
NaiqiLi, WenjieLi, JifengSun....
Published date-12/01/2020
GaussianProcesses, VariationalInference
In this paper we propose Stochastic Deep Gaussian Processes over Graphs (DGPG), which are deep structure models that learn the mappings between input and output signals in graph domains. The …