Contrastive Energy

Contrastive energy methods leverage the difference in energy levels between similar and dissimilar data points to learn effective representations or models. Current research focuses on applying these methods to diverse tasks, including classification, reinforcement learning, and domain generalization, often employing energy-based models and contrastive learning frameworks within deep neural networks. This approach shows promise in improving the robustness and efficiency of various machine learning algorithms, particularly in high-dimensional data settings, and offers advantages in interpretability and generative capabilities. The resulting advancements have implications for various fields, from medical image analysis to improving the performance of AI systems.

Papers

June 24, 2023

Robust Classification of High-Dimensional Data using Data-Adaptive Energy Distance
Jyotishka Ray Choudhury, Aytijhya Saha, Sarbojit Roy, Subhajit Dutta
Simple Classifier High Dimensional Data Classification Algorithm Robust Classification Contrastive Energy Low Sample Size

April 25, 2023

Contrastive Energy Prediction for Exact Energy-Guided Diffusion Sampling in Offline Reinforcement Learning
Cheng Lu, Huayu Chen, Jianfei Chen, Hang Su, Chongxuan Li, Jun Zhu
Reinforcement Learning Diffusion Model Offline Reinforcement Learning Diffusion Sampling Intermediate Language Contrastive Energy Guided Sampling

April 7, 2023

Conservative objective models are a special kind of contrastive divergence-based energy model
Christopher Beckham, Christopher Pal
Gradient Based Different Type Offline Model Based Optimization Contrastive Energy Conservative Objective Model

February 22, 2023

Energy-Based Test Sample Adaptation for Domain Generalization
Zehao Xiao, Xiantong Zhen, Shengcai Liao, Cees G. M. Snoek
Domain Generalization Test Time Adaptation Trained Source Model Contrastive Energy Sample Efficient Adaptation

March 25, 2022

A Unified Contrastive Energy-based Model for Understanding the Generative Ability of Adversarial Training
Yifei Wang, Yisen Wang, Jiansheng Yang, Zhouchen Lin
Generative Model Adversarial Training Adversarial Learning Unified Alignment Adversarial Data Collection Generative Power Contrastive Energy

March 7, 2022

Covariate-Balancing-Aware Interpretable Deep Learning models for Treatment Effect Estimation
Kan Chen, Qishuo Yin, Qi Long
Treatment Effect Outcome Prediction Interpretable Deep Learning Average Treatment Effect Contrastive Energy

February 10, 2022

Energy-Based Contrastive Learning of Visual Representations
Beomsu Kim, Jong Chul Ye
Contrastive Learning Visual Representation Generative Learning Contrastive Energy

Contrastive Energy

Papers

Robust Classification of High-Dimensional Data using Data-Adaptive Energy Distance

Contrastive Energy Prediction for Exact Energy-Guided Diffusion Sampling in Offline Reinforcement Learning

Conservative objective models are a special kind of contrastive divergence-based energy model

Energy-Based Test Sample Adaptation for Domain Generalization

A Unified Contrastive Energy-based Model for Understanding the Generative Ability of Adversarial Training

Covariate-Balancing-Aware Interpretable Deep Learning models for Treatment Effect Estimation

Energy-Based Contrastive Learning of Visual Representations