Marc Finzi

Papers

1

Total Citations

14

H-Index

1

About

Marc Finzi is a leading researcher at the intersection of machine learning and physics, with a focus on designing neural networks that incorporate powerful inductive biases from the natural sciences. His most cited work, "Deconstructing the Inductive Biases of Hamiltonian Neural Networks" (2022, 14 citations), critically examines physics-inspired models like Hamiltonian and Lagrangian neural networks. While these models can dramatically outperform standard learned dynamics by enforcing energy conservation, Finzi highlights their limitations in real-world systems that do not conserve energy, such as those with friction or external forces. This contribution has been pivotal in guiding the field toward more flexible, yet principled, approaches to modeling physical systems. Beyond this, Finzi has made notable strides in scalable machine learning, including work on equivariant neural networks and optimization algorithms. His research is widely recognized for bridging theoretical rigor with practical applicability, earning him citations and acclaim among both physicists and AI researchers.

Research Focus

Key Achievements

1
H-Index
1
Papers
14
Total Citations
14
Avg Citations/Paper
🏆 Most Cited Paper
Deconstructing the Inductive Biases of Hamiltonian Neural Networks
14 citations · 2022
📈 Most Prolific Year: 2022 (1 Papers)
🤝 Key Collaborators: 3

Top Papers

  1. 1

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 9 days ago