Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Adaptive Posterior Learning: few-shot learning with a surprise-based memory module
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 |
|---|---|---|---|
| Use surprise-based selective memory writing | APL · Few-shot classification | memory footprint: roughly 8x memory cost | Helped |
| Set the surprise threshold σ to 0 and write every example (ablation) | APL · 200-way classification | accuracy: slightly higher | Helped |
| Set the surprise threshold σ to 0 and write every example (ablation) | APL · 200-way classification | memory size: roughly 8x memory cost | Hurt |
| Vary the surprise threshold σ over a large range (ablation) | APL · 200-way classification | accuracy | No clear effect |
| Vary the surprise threshold σ over a large range (ablation) | APL · 200-way classification | memory size | No clear effect |
| Replace the relational self-attention feed-forward decoder with relational working memory or LSTM (ablation) | APL · Omniglot classification | accuracy | No clear effect |
| Use the relational self-attention feed-forward decoder rather than the other decoder architectures (ablation) | APL · Number analogy | accuracy | Helped |
| Retrieve k=8 or k=16 neighbors for relational reasoning (ablation) | APL · Number analogy with shuffled numbers and symbols | accuracy | Helped |
Closest papers by meaning
- Addressing Catastrophic Forgetting in Few-Shot Problems (ICML 2021 · similarity 0.87)
- Learning to Learn Variational Semantic Memory (NeurIPS 2020 · similarity 0.86)
- Memory Efficient Online Meta Learning (ICML 2021 · similarity 0.85)
- Learning to Forget for Meta-Learning (CVPR 2020 · similarity 0.85)
- Efficient Adaptive Data Acquisition via Pretrained Belief Representations (NeurIPS 2026 · similarity 0.85)
- The Wasserstein Believer: Learning Belief Updates for Partially Observable Environments through Reliable Latent Space Models (ICLR 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.