TODO.md

Roadmap

Solver

• [] Conjugate Gradient
• [] GMRES
• [] Type-I/II Anderson
• [] Broyden class
• [] HyperSolver

Backward Pass

• [] SHINE
  • SHINE: SHaring the INverse Estimate from the forward pass for bi-level optimization and implicit models

DEQ Zoo

• [] monDEQ
  • Monotone operator equilibrium networks
• [] DE-Prox
  • Deep Equilibrium Architectures for Inverse Problems in Imaging
• [] JIIO
  • Joint inference and input optimization in equilibrium networks
• [] EIGNN
  • EIGNN: Efficient Infinite-Depth Graph Neural Networks
• [] MOptEq
  • Optimization inspired Multi-Branch Equilibrium Models
• [] GIND
  • Optimization-Induced Graph Implicit Nonlinear Diffusion
• [] Adversarial Defense
  • A Closer Look at the Adversarial Robustness of Deep Equilibrium Models
• [] MGNNI
  • MGNNI: Multiscale Graph Neural Networks with Implicit Layers
• [] Path Independence
  • Path Independent Equilibrium Models Can Better Exploit Test-Time Computation
• [] NERD
  • Equilibrium Image Denoising With Implicit Differentiation
• [] Entropy Reduction
  • Improving Adversarial Robustness of Deep Equilibrium Models with Explicit Regulations Along the Neural Dynamics
• [] DEQNAR
  • Deep Equilibrium Non-Autoregressive Sequence Learning
• [] LDEQ
  • Recurrence without Recurrence: Stable Video Landmark Detection with Deep Equilibrium Models
• [] DEQDet
  • Deep Equilibrium Object Detection
• [] FNO-DEQ
  • Deep Equilibrium Based Neural Operators for Steady-State PDEs
• [] GET
  • One-Step Diffusion Distillation via Deep Equilibrium Models

HuggingFace

To be continues.