Research map / 3D vision / Face and shape
Einstein Fields: A Neural Perspective To Computational General Relativity
Sandeep Suresh Cranganore, Andrei Bodnar, Arturs Berzins, Johannes Brandstetter
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| model metric distortion instead of full metric (ablation) | EinFields · Schwarzschild ablation | Rel. l2: 2.13e-6 -> 1.40e-7 | Improved |
| use cosine LR schedule instead of constant LR (ablation) | EinFields · Schwarzschild ablation | Rel. l2: 2.37e-5 -> 1.40e-7 | Improved |
| use SOAP optimizer instead of ADAM (ablation) | EinFields · Schwarzschild ablation | Rel. l2: 4.16e-6 -> 1.40e-7 | Improved |
| use SiLU activation instead of WIRE (ablation) | EinFields · Schwarzschild ablation | Rel. l2: 4.12e-6 -> 1.40e-7 | Improved |
| add Hessian supervision on top of Jacobian (ablation) | EinFields · Schwarzschild ablation | Rel. l2: 1.51e-7 -> 1.40e-7 | Improved |
| add Jacobian and Hessian supervision (Sobolev training) (ablation) | EinFields · Schwarzschild ablation | Rel. l2: 2.37e-7 -> 1.40e-7 | Improved |
| add Jacobian supervision | EinFields · Schwarzschild metric | Rel. l2: 1.08e-6 -> 3.37e-7 | Improved |
| add Jacobian and Hessian supervision | EinFields · Schwarzschild metric | Rel. l2: 1.08e-6 -> 1.88e-7 | Improved |
| add Jacobian and Hessian supervision | EinFields · Schwarzschild metric | compression factor: 4035 -> 1698 | Hurt |
| add Jacobian supervision | EinFields · Neutron star NR simulation | Rel. l2: 3.60e-5 -> 6.95e-6 | Improved |
| use EinFields AD instead of FD stencils (h=0.01) for Riemann tensor | EinFields · Schwarzschild Riemann tensor | MAE: 1.78e-2 -> 1.25e-6 | Improved |
| use EinFields AD instead of FD stencils for Christoffel symbols | EinFields · Schwarzschild Christoffel symbol | MAE: 5.37e-6 -> 9.98e-7 | Improved |
Closest papers by meaning
- Riemann Tensor Neural Networks: Learning Conservative Systems with Physics-Constrained Networks (ICML 2025 · similarity 0.84)
- Neural Spacetimes for DAG Representation Learning (ICLR 2025 · similarity 0.84)
- Clifford-Steerable Convolutional Neural Networks (ICML 2024 · similarity 0.82)
- Lorentz Local Canonicalization: How to make any Network Lorentz-Equivariant (NeurIPS 2025 · similarity 0.82)
- How to Train Neural Field Representations: A Comprehensive Study and Benchmark (CVPR 2024 · similarity 0.82)
- Universal Physics Transformers: A Framework For Efficiently Scaling Neural Operators (NeurIPS 2024 · similarity 0.82)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
