Research map / Label-efficient and robust learning / Domain adaptation and generalization
Transfer Learning for Sequences via Learning to Collocate
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 |
|---|---|---|---|
| Add cell-level transfer with attention over collocated source-domain positions, compared with corresponding-position transfer only (ablation) | ART · Amazon reviews (average) | accuracy: 0.803 -> 0.858 | Helped |
| Add cell-level transfer rather than transferring only the final RNN cell (ablation) | ART · Amazon reviews (average) | accuracy: 0.774 -> 0.858 | Helped |
| Use ART instead of a target-only LSTM | Amazon reviews (average) · Amazon reviews | accuracy: 0.763 -> 0.858 | Helped |
| Use ART instead of the strongest reported baseline | ART · Amazon reviews, scarce target training (average) | accuracy: 0.856 -> 0.876 | Helped |
| Use ART instead of HRN | ART · POS tagging, PTB → Twitter/0.1 | per-word accuracy: 0.837 -> 0.859 | Helped |
| Use ART instead of CCT | ART · NER, CoNLL → Twitter/0.1 | F1: 0.434 -> 0.450 | Helped |
| Use ART instead of CCT | ART · NER, Twitter → CoNLL/0.01 | F1: 0.675 -> 0.707 | Helped |
| Use ART instead of FLORS | ART · POS tagging, PTB → Twitter/0.01 | per-word accuracy: 0.763 -> 0.658 | Hurt |
Closest papers by meaning
- Transfer Learning for Sequence Tagging with Hierarchical Recurrent Networks (ICLR 2017 · similarity 0.89)
- Transfer Learning for Sequence Labeling Using Source Model and Target Data (AAAI 2019 · similarity 0.88)
- Empower Sequence Labeling with Task-Aware Neural Language Model (AAAI 2018 · similarity 0.86)
- RNN-Based Sequence-Preserved Attention for Dependency Parsing (AAAI 2018 · similarity 0.85)
- Recurrent Stacking of Layers for Compact Neural Machine Translation Models (AAAI 2019 · similarity 0.85)
- GIRNet: Interleaved Multi-Task Recurrent State Sequence Models (AAAI 2019 · 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.