Research map / Learning theory and algorithms / Prediction and decision losses
Gone Fishing: Neural Active Learning with Fisher Embeddings
Jordan T. Ash, Surbhi Goel, Akshay Krishnamurthy, Sham M. Kakade
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 the full-data Fisher term I(theta) to the trace objective (BAIT vs BAIT without it) (ablation) | BAIT · MNIST linear classification, random projection | accuracy | Improved |
| use trace objective with full-rank Fisher instead of determinant maximization (ablation) | greedy batch selection · MNIST linear classification | accuracy | Improved |
| add backward pruning step to forward greedy selection (ablation) | BAIT · MNIST MLP, batch 10 | accuracy | Mixed |
| increase oversampling factor beyond 2 (ablation) | BAIT · active learning | accuracy | No clear effect |
| use BAIT instead of BADGE/Coreset/Confidence/Random | deep classification (MNIST, SVHN, CIFAR-10, OpenML 155) · pairwise penalty matrix | pairwise significant wins | Improved |
| use BAIT instead of Coreset/Random | active regression (yearbook, rainfall, MNIST, SVHN) · regression learning curves | error | Improved |
Closest papers by meaning
- Neural Active Learning Beyond Bandits (ICLR 2024 · similarity 0.87)
- Towards Robust and Reproducible Active Learning Using Neural Networks (CVPR 2022 · similarity 0.87)
- Fishers for Free? Approximating the Fisher Information Matrix by Recycling the Squared Gradient Accumulator (ICML 2025 · similarity 0.86)
- Provably Neural Active Learning Succeeds via Prioritizing Perplexing Samples (ICML 2024 · similarity 0.85)
- Transductive Active Learning: Theory and Applications (NeurIPS 2024 · similarity 0.85)
- Active Learning Guided by Efficient Surrogate Learners (AAAI 2024 · similarity 0.85)
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.
