Research map / Architectures and efficiency / RNNs and normalization
Stochastic Deep Networks with Linear Competing Units for Model-Agnostic Meta-Learning
Konstantinos Kalais, Sotirios Chatzis
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 stochastic instead of deterministic LWTA competition (ablation) | StochLWTA-ML · Omniglot 20-way 1-shot | accuracy: 96.95 -> 97.79 | Improved |
| Use stochastic instead of deterministic LWTA competition (ablation) | StochLWTA-ML · Omniglot 20-way 5-shot | accuracy: 98.63 -> 98.97 | Improved |
| Use stochastic instead of deterministic LWTA competition (ablation) | StochLWTA-ML · Mini-Imagenet 5-way 1-shot | accuracy: 53.12 -> 54.11 | Improved |
| Use stochastic instead of deterministic LWTA competition (ablation) | StochLWTA-ML · Mini-Imagenet 5-way 5-shot | accuracy: 64.93 -> 66.70 | Improved |
| Increase competing units per block from J=2 to J=4 (ablation) | StochLWTA-ML · Omniglot 20-way 1-shot | accuracy: 97.79 -> 96.33 | Hurt |
| Increase competing units per block from J=2 to J=8 (ablation) | StochLWTA-ML · Omniglot 20-way 1-shot | accuracy: 97.79 -> 95.38 | Hurt |
| Replace Gaussian weights with point-estimate weights (ablation) | StochLWTA-ML · Omniglot 20-way 1-shot | accuracy: 97.68 -> 97.79 | Hurt |
| Replace Gaussian weights with point-estimate weights (ablation) | StochLWTA-ML · Omniglot 20-way 5-shot | accuracy: 98.85 -> 98.97 | Hurt |
| Use a stochastic LWTA network with 4 convolutional layers instead of the compact architecture (ablation) | StochLWTA-ML · Omniglot 20-way 1-shot | accuracy: 97.79 -> 96.50 | Hurt |
| Use stochastic LWTA and Bayesian weights instead of deterministic ReLU networks with point-estimate weights (ablation) | Meta-learning classifiers · Omniglot 20-way 1-shot | accuracy: 2.2% and 3.5% reduction | Improved |
Closest papers by meaning
- Competing Mutual Information Constraints with Stochastic Competition-Based Activations for Learning Diversified Representations (AAAI 2022 · similarity 0.89)
- Simple and Effective Stochastic Neural Networks (AAAI 2021 · similarity 0.85)
- Partially Stochastic Infinitely Deep Bayesian Neural Networks (ICML 2024 · similarity 0.84)
- Random Sparse Lifts: Construction, Analysis and Convergence of finite sparse networks (ICLR 2024 · similarity 0.84)
- How Deep Networks Learn Sparse and Hierarchical Data: the Sparse Random Hierarchy Model (ICML 2024 · similarity 0.83)
- Efficient Training of Neural Stochastic Differential Equations by Matching Finite Dimensional Distributions (ICLR 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.
