Research map / Architectures and efficiency / Neural architecture search
Transfer Learning with Neural AutoML
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 |
|---|---|---|---|
| Transfer a multitask-pretrained controller and learn a new task embedding | Neural AutoML · 13 NLP classification tasks | trials to match Random Search validation accuracy-top10: In 12 of the 13 datasets T-AutoML achieves the desired reward fastest; in 9 cases achieves an order of magnitude speed-u | Helped |
| Remove task embeddings from the transferred multitask controller (ablation) | T-NAML · 13 NLP classification tasks | test accuracy-top10: mean performance drop ... 1.8\% | Mixed |
| Transfer a single architecture instead of the controller parameters (ablation) | T-NAML · 13 NLP classification tasks | test accuracy-top10: mean performance increase of 4.8\% | Hurt |
| Pretrain the controller jointly on multiple tasks | Multitask Neural AutoML · 8 NLP tasks | accuracy | No clear effect |
| Transfer from English-task priors to a Spanish-language task | T-NAML · Sentiment Cine | test accuracy-top10: 79.8\% ... 79.4\% ... 78.1\% | Mixed |
Closest papers by meaning
- AutoMMLab: Automatically Generating Deployable Models from Language Instructions for Computer Vision Tasks (AAAI 2025 · similarity 0.85)
- AutoML-Agent: A Multi-Agent LLM Framework for Full-Pipeline AutoML (ICML 2025 · similarity 0.84)
- AutoAL: Automated Active Learning with Differentiable Query Strategy Search (ICML 2025 · similarity 0.83)
- MLZero: A Multi-Agent System for End-to-end Machine Learning Automation (NeurIPS 2025 · similarity 0.83)
- UniPT: Universal Parallel Tuning for Transfer Learning with Efficient Parameter and Memory (CVPR 2024 · similarity 0.83)
- Building Optimal Neural Architectures using Interpretable Knowledge (CVPR 2024 · 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.