Latent Embeddings

Latent embeddings are low-dimensional representations of data that capture essential information while reducing dimensionality, aiming to improve efficiency and performance in various machine learning tasks. Current research focuses on developing methods to generate these embeddings effectively, particularly using techniques like variational autoencoders, transformers, and graph neural networks, often within self-supervised or semi-supervised learning frameworks. These advancements are impacting diverse fields, including recommendation systems, image analysis, natural language processing, and even healthcare, by enabling more efficient data processing, improved model generalization, and enhanced interpretability. The ability to learn robust and meaningful latent embeddings is crucial for advancing many areas of artificial intelligence.

Papers

October 14, 2022

Vision Transformer Visualization: What Neurons Tell and How Neurons Behave?
Van-Anh Nguyen, Khanh Pham Dinh, Long Tung Vuong, Thanh-Toan Do, Quan Hung Tran, Dinh Phung, Trung Le
Vision Transformer Computer Vision CNN Model Single Neuron Level Latent Embeddings

September 14, 2022

Predicting Drug-Drug Interactions using Deep Generative Models on Graphs
Nhat Khang Ngo, Truong Son Hy, Risi Kondor
Graph Drawing Latent Space Latent Representation Variational Graph Drug Drug Interaction Latent Embeddings

May 20, 2022

Learning Geometrically Disentangled Representations of Protein Folding Simulations
N. Joseph Tatro, Payel Das, Pin-Yu Chen, Vijil Chenthamarakshan, Rongjie Lai
Molecular Dynamic Latent Embeddings Conformational Ensemble Biomolecular Simulation Pre Trained Protein Model Protein Backbone Generation

April 1, 2022

Learnable latent embeddings for joint behavioral and neural analysis
Steffen Schneider, Jin Hwa Lee, Mackenzie Weygandt Mathis
Neural Representation Neural Activity Neural Dynamic Latent Embeddings Complex Behavior Neural Metric

February 18, 2022

Explaining, Evaluating and Enhancing Neural Networks' Learned Representations
Marco Bertolini, Djork-Arné Clevert, Floriane Montanari
Neural Network Transfer Learning Representation Learning Inherent Interpretability Line by Line Explanation High Explainability Meaningful Representation Latent Embeddings

December 9, 2021

Few-Shot Keypoint Detection as Task Adaptation via Latent Embeddings
Mel Vecerik, Jackie Kay, Raia Hadsell, Lourdes Agapito, Jon Scholz
Computer Vision Task Sparse Network Keypoint Detection Task Adaptation Shot Adaptation Latent Embeddings Dense Embeddings