Research map / Vision-language and multimodal / Visual question answering
Dynamic Coattention Networks For Question Answering
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 a DCN ensemble instead of the strongest listed ensemble baseline | DCN · SQuAD test | F1: 78.3 -> 80.4 | Helped |
| Use DCN instead of the strongest listed single-model baseline | single-model QA · SQuAD test | F1: 75.0 -> 75.9 | Helped |
| Add an iterative decoder rather than forcing a single iteration (ablation) | DCN · SQuAD dev | F1: 74.0 -> 75.6 | Helped |
| Use HMN instead of a 2-layer MLP (ablation) | DCN decoder · SQuAD dev | F1: 74.4 -> 75.6 | Helped |
| Use coattention instead of the cited attention mechanism (ablation) | DCN encoder · SQuAD dev | F1: 73.7 -> 75.6 | Helped |
| Use HMN maxout pool size 16 rather than 8 (ablation) | DCN decoder · SQuAD dev | F1: 74.9 -> 75.6 | Helped |
| Use HMN maxout pool size 16 rather than 4 (ablation) | DCN decoder · SQuAD dev | F1: 75.2 -> 75.6 | Helped |
| Add an attention mechanism rather than removing attention (ablation) | DCN · SQuAD | F1 | Helped |
Closest papers by meaning
- DCN+: Mixed Objective And Deep Residual Coattention for Question Answering (ICLR 2018 · similarity 0.91)
- Dynamic Memory Networks for Visual and Textual Question Answering (ICML 2016 · similarity 0.89)
- Ask Me Anything: Dynamic Memory Networks for Natural Language Processing (ICML 2016 · similarity 0.88)
- Coarse-grain Fine-grain Coattention Network for Multi-evidence Question Answering (ICLR 2019 · similarity 0.88)
- Beyond RNNs: Positional Self-Attention with Co-Attention for Video Question Answering (AAAI 2019 · similarity 0.87)
- GLEN: Generalized Focal Loss Ensemble of Low-Rank Networks for Calibrated Visual Question Answering (AAAI 2025 · 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.