96 citations · 111 across the 9 of their papers we have counts for
9 papers
Exact Global MCMC with Denoising Diffusion
Mitch Hill
This work shows that diffusion models learned with standard denoising loss can provide effective global MCMC proposals for complex high-dimensional target densities. The method is…
Memory Efficient Full-gradient Attacks (MEFA) Framework for Adversarial Defense Evaluations
Yuan Du, Mitchel Hill, HanQin Cai
This work studies the robust evaluation of iterative stochastic purification defenses under white-box adversarial attacks. Our key technical insight is that gradient checkpointing…
Learning Probabilistic Models from Generator Latent Spaces with Hat EBM
Mitch Hill, Erik Nijkamp, Jonathan Mitchell +2
This work proposes a method for using any generator network as the foundation of an Energy-Based Model (EBM). Our formulation posits that observed images are the sum of unobserved…
EBM Life Cycle: MCMC Strategies for Synthesis, Defense, and Density Modeling
Mitch Hill, Jonathan Mitchell, Chu Chen +3
This work presents strategies to learn an Energy-Based Model (EBM) according to the desired length of its MCMC sampling trajectories. MCMC trajectories of different lengths corresp…
Stochastic Security: Adversarial Defense Using Long-Run Dynamics of Energy-Based Models
Mitch Hill, Jonathan Mitchell, Song-Chun Zhu
The vulnerability of deep networks to adversarial attacks is a central problem for deep learning from the perspective of both cognition and security. The current most successful de…
Learning Non-Convergent Non-Persistent Short-Run MCMC Toward Energy-Based Model
Erik Nijkamp, Mitch Hill, Song-Chun Zhu +1
This paper studies a curious phenomenon in learning energy-based model (EBM) using MCMC. In each learning iteration, we generate synthesized examples by running a non-convergent, n…