MEDALLION: Manifold Embedding Distillation and Attributions for Low-dimensional Latent InterpretatiON
cs.LG, stat.ML
Submitted: 2026-11-05
Updated: 2026-11-05
Terminology
Sources
- Interpretable Machine Learning for Discovery: Statistical Challenges \& Opportunities
- Local Explanation of Dimensionality Reduction
- Interpretability via Model Extraction
- Low-dimensional embeddings of high-dimensional data
- MEDAL: Manifold Embedding Distillation via Autoencoder Learning
- Improving Knowledge Distillation with Teacher's Explanation
- Distilling a Neural Network Into a Soft Decision Tree
- On the Impact of Knowledge Distillation for Model Interpretability
- Distilling the Knowledge in a Neural Network
- Stop Misusing t-SNE and UMAP for Visual Analytics
- Calibrating dimension reduction hyperparameters in the presence of noise
- UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction
- Good Teachers Explain: Explanation-Enhanced Knowledge Distillation
- Learning Important Features Through Propagating Activation Differences
- SmoothGrad: removing noise by adding noise
- Axiomatic Attribution for Deep Networks
- Unsupervised Learning of Neural Networks to Explain Neural Networks
- Approximation Trees: Statistical Stability in Model Distillation
Related papers
- A Physics-Regulated Neural Framework for Learning 3D Grain Growth Dynamics
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis