Research map / Learning theory and algorithms / Sample complexity
Improved Active Multi-Task Representation Learning via Lasso
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 |
|---|---|---|---|
| Replace L2-regularized relevance (L2-A-MTRL) sampling with L1-regularized relevance (L1-A-MTRL) sampling | active multi-task representation learning on MNIST-C · MNIST-C full task scenario (linear head, regression loss) | average accuracy: +0.54% | Helped |
| Restrict sampling to at most 10 source tasks (sparse k-task selection) with L1-A-MTRL instead of uniform sampling | L1-A-MTRL on MNIST-C · MNIST-C k-task selection scenario (ConvNet representation) | average accuracy: +2.2% over uniform sampling | Helped |
Closest papers by meaning
- Active Multi-Task Representation Learning (ICML 2022 · similarity 0.92)
- Active representation learning for general task space with applications in robotics (NeurIPS 2023 · similarity 0.88)
- Near-Optimal and Efficient First-Order Algorithm for Multi-Task Learning with Shared Linear Representation (ICML 2026 · similarity 0.87)
- Provable Multi-Task Representation Learning by Two-Layer ReLU Neural Networks (ICML 2024 · similarity 0.86)
- Learning Multi-Task Sparse Representation Based on Fisher Information (AAAI 2024 · similarity 0.84)
- Joint-Task Regularization for Partially Labeled Multi-Task Learning (CVPR 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.