Research map / Label-efficient and robust learning / Recognition and long-tail
NEU: A Meta-Algorithm for Universal UAP-Invariant Feature Representation
Anastasis Kratsios, Cody Hyndman
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 |
|---|---|---|---|
| add NEU feature map to the DNN | DNN · Apple stock tracker, test | MSE: 5.112289 -> 4.948565 | Improved |
| add NEU feature map to elastic net | ENET · Apple stock tracker, test | MSE: 4.903049 -> 5.369902 | Hurt |
| add NEU feature map to kernel ridge | kRidge · Apple stock tracker, test | MSE: 5.110043 -> 5.515611 | Hurt |
| add NEU feature map to gradient boosted random forests | GBRF · Apple stock tracker, test | MSE: 5.971403 -> 5.989319 | No clear effect |
| add NEU feature map to the auto-encoder (1 factor) | AE · US bond yield curve, test | MSE: 14.792091 -> 11.043391 | Improved |
| add NEU feature map to the auto-encoder (2 factors) | AE · US bond yield curve, test | MSE: 11.930328 -> 10.425360 | Improved |
| add NEU feature map to the auto-encoder (3 factors) | AE · US bond yield curve, test | MSE: 14.688698 -> 13.579660 | Improved |
| add NEU feature map to PCA (3 factors) | PCA · US bond yield curve, test | MSE: 15.176116 -> 14.439484 | Improved |
| add NEU feature map to PCA (1 factor) | PCA · US bond yield curve, test | MSE: 15.302093 -> 15.518402 | No clear effect |
| add NEU feature map to kernel PCA (2 factors) | kPCA · US bond yield curve, test | MSE: 15.614535 -> 21.080980 | Hurt |
| add NEU feature map to kernel ridge | kRidge · Simulation, jump discontinuity, test | MSE: 0.088674 -> 0.074313 | Improved |
| add NEU feature map to DNN | DNN · Simulation, aperiodic oscillations, test | MSE: 0.304498 -> 0.296543 | Improved |
Closest papers by meaning
- Unsupervised Feature Learning with Emergent Data-Driven Prototypicality (CVPR 2024 · similarity 0.84)
- On Feature Learning in Structured State Space Models (NeurIPS 2024 · similarity 0.83)
- URLOST: Unsupervised Representation Learning without Stationarity or Topology (ICLR 2025 · similarity 0.83)
- Finding NEM-U: Explaining unsupervised representation learning through neural network generated explanation masks (ICML 2024 · similarity 0.83)
- FeatUp: A Model-Agnostic Framework for Features at Any Resolution (ICLR 2024 · similarity 0.83)
- Unified Gradient-Based Machine Unlearning with Remain Geometry Enhancement (NeurIPS 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.
