Research map / Learning theory and algorithms / Combinatorial and search optimization
Efficient and Modular Implicit Differentiation
Mathieu Blondel, Quentin Berthet, marco cuturi, Roy Frostig, Stephan Hoyer, Felipe Llinares-López, Fabian Pedregosa, Jean-Philippe Vert
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 |
|---|---|---|---|
| use implicit differentiation instead of differentiating unrolled iterates for the inner problem | dataset distillation with gradient descent solver · MNIST dataset distillation | runtime: 4 times faster | Improved |
| train dictionary with a task-driven outer logistic loss rather than reconstruction loss only | dictionary learning + L2 logreg · Breast cancer survival prediction | AUC (%): 68.3 -> 73.2 | Improved |
| task-driven dictionary learning vs L1/L2 logistic regression with 100x fewer variables | Task-driven DictL · Breast cancer survival prediction | AUC (%): 72.4 -> 73.2 | No clear effect |
| differentiate unrolled optimizer instead of implicit differentiation | molecular dynamics position sensitivity · Molecular dynamics sensitivity | convergence | Hurt |
Closest papers by meaning
- Automatic Differentiation in Machine Learning: a Survey (JMLR 2018 · similarity 0.88)
- SoftJAX & SoftTorch: Empowering Automatic Differentiation Libraries with Informative Gradients (ICML 2026 · similarity 0.86)
- Nonsmooth Implicit Differentiation: Deterministic and Stochastic Convergence Rates (ICML 2024 · similarity 0.86)
- One-step differentiation of iterative algorithms (NeurIPS 2023 · similarity 0.85)
- Collapsing Taylor Mode Automatic Differentiation (NeurIPS 2025 · similarity 0.83)
- An Interpretable Approach to the Solutions of High-Dimensional Partial Differential Equations (AAAI 2024 · similarity 0.83)
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.
