Research map / Optimization / Nonconvex and smooth optimization
Interpretable Neural Networks with Frank-Wolfe: Sparse Relevance Maps and Relevance Orderings
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Combine rate-constrained RDE solutions across selected rates | RDE · MNIST | distortion | Helped |
| Optimize the relevance ordering across rates directly | RDE · MNIST | distortion | Helped |
| Optimize the relevance ordering across rates directly | RDE · STL-10 | distortion | Helped |
| Use Frank-Wolfe instead of PGD (ablation) | RC-RDE · MNIST | distortion | Helped |
| Use Frank-Wolfe instead of PGD (ablation) | RC-RDE · STL-10, rates up to 40% | distortion | Helped |
| Use Frank-Wolfe instead of PGD (ablation) | RC-RDE · STL-10, rates above 40% | distortion | No clear effect |
| Use any of the tested Frank-Wolfe variants (ablation) | RC-RDE · MNIST | distortion | No clear effect |
| Use SFW rather than deterministic FW variants (ablation) | Ord-RDE · MNIST, rates between 20% and 80% | distortion | Helped |
| Use SFW rather than deterministic FW variants (ablation) | Ord-RDE · MNIST, very low and high rates | distortion | No clear effect |
| Increase the SFW batch size over iterations (ablation) | Ord-RDE · MNIST | distortion | No clear effect |
| Add momentum to SFW (ablation) | Ord-RDE · MNIST | distortion | Hurt |
Closest papers by meaning
- Explainable Neural Networks with Guarantee: A Sparse Estimation Approach (AAAI 2025 · similarity 0.88)
- Explaining, Fast and Slow: Abstraction and Refinement of Provable Explanations (ICML 2025 · similarity 0.84)
- Predictive Concept Decoders: Training Scalable End-to-End Interpretability Assistants (NeurIPS 2026 · similarity 0.83)
- Q-SENN: Quantized Self-Explaining Neural Networks (AAAI 2024 · similarity 0.83)
- Enhancing Performance of Explainable AI Models with Constrained Concept Refinement (ICML 2025 · similarity 0.83)
- Concept-Guided Interpretability via Neural Chunking (NeurIPS 2025 · 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.