Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Meta-Neighborhoods
Siyuan Shan, Yang Li, Junier B Oliva
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 Meta-Neighborhoods to the parametric classifier | 4-layer ConvNet · SVHN | accuracy: 93.02±0.12% -> 94.46±0.09% | Improved |
| Add iFiLM alongside Meta-Neighborhoods (ablation) | Meta-Neighborhoods with 4-layer ConvNet · SVHN | accuracy: 94.46±0.09% -> 94.92±0.09% | Improved |
| Add Meta-Neighborhoods to the parametric classifier | DenseNet40-BC · CIFAR-100 | accuracy: 73.92±0.12% -> 76.32±0.16% | Improved |
| Add iFiLM alongside Meta-Neighborhoods (ablation) | Meta-Neighborhoods with DenseNet40-BC · CIFAR-100 | accuracy: 76.32±0.16% -> 76.96±0.14% | Improved |
| Add Meta-Neighborhoods to the parametric classifier | ResNet-56 · ImageNet | accuracy: 48.41±0.14% -> 51.85±0.12% | Improved |
| Add iFiLM alongside Meta-Neighborhoods (ablation) | Meta-Neighborhoods with ResNet-56 · ImageNet | accuracy: 51.85±0.12% -> 54.23±0.13% | Improved |
| Use Meta-Neighborhoods rather than the vanilla regressor | MLP · music | MSE: 0.6236±0.0056 -> 0.6088±0.0050 | Improved |
| Use Meta-Neighborhoods rather than the vanilla regressor | MLP · toms | MSE: 0.0594±0.0080 -> 0.0531±0.0073 | Improved |
| Use Meta-Neighborhoods rather than the vanilla regressor | MLP · cte | MSE: 0.00121±0.00022 -> 0.00109±0.00015 | Improved |
| Use Meta-Neighborhoods rather than the vanilla regressor | MLP · super | MSE: 0.1132±0.0060 -> 0.1077±0.0068 | Improved |
| Use Meta-Neighborhoods rather than the vanilla regressor | MLP · gom | MSE: 0.5949±0.0515 -> 0.5681±0.0563 | Improved |
| Use a scalar inner-loop learning rate rather than a diagonal matrix (ablation) | Meta-Neighborhoods · CIFAR-100, DenseNet40-BC | accuracy: 76.32% -> 77.20% | Hurt |
Closest papers by meaning
- Meta Learning in Decentralized Neural Networks: Towards More General AI (AAAI 2023 · similarity 0.86)
- Meta-Learning in Neural Networks: A Survey (TPAMI 2021 · similarity 0.86)
- $\mu$LO: Compute-Efficient Meta-Generalization of Learned Optimizers (ICLR 2026 · similarity 0.85)
- On Modulating the Gradient for Meta-Learning (ECCV 2020 · similarity 0.85)
- Meta-Learning and Universality: Deep Representations and Gradient Descent can Approximate any Learning Algorithm (ICLR 2018 · similarity 0.85)
- On the Stability and Generalization of Meta-Learning (NeurIPS 2024 · similarity 0.84)
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.
