Research map / Architectures and efficiency / RNNs and normalization
A Compare-Aggregate Model for Matching Text Sequences
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 SubMult+NN comparison rather than NN (ablation) | Compare-aggregate model · MovieQA dev | accuracy: 31.6 -> 72.1 | Helped |
| Use SubMult+NN comparison rather than EucCos (ablation) | Compare-aggregate model · MovieQA dev | accuracy: 71.9 -> 72.1 | No clear effect |
| Use SubMult+NN comparison rather than Mult (ablation) | Compare-aggregate model · InsuranceQA test1 | accuracy: 75.2 -> 75.6 | Helped |
| Use SubMult+NN comparison rather than Mult (ablation) | Compare-aggregate model · InsuranceQA test2 | accuracy: 73.4 -> 72.3 | Hurt |
| Use element-wise Mult rather than SubMult+NN (ablation) | Compare-aggregate model · WikiQA | MAP: 0.7332 -> 0.7433 | Helped |
| Use element-wise Mult rather than SubMult+NN (ablation) | Compare-aggregate model · WikiQA | MRR: 0.7477 -> 0.7545 | Helped |
| Use element-wise Sub rather than SubMult+NN (ablation) | Compare-aggregate model · SNLI test | accuracy: 86.8 -> 86.8 | No clear effect |
| Use SubMult+NN comparison rather than EucCos (ablation) | Compare-aggregate model · InsuranceQA test1 | accuracy: 70.2 -> 75.6 | Helped |
Closest papers by meaning
- Deconvolutional Latent-Variable Model for Text Sequence Matching (AAAI 2018 · similarity 0.87)
- Multi-Level Head-Wise Match and Aggregation in Transformer for Textual Sequence Matching (AAAI 2020 · similarity 0.86)
- Text Alignment Is An Efficient Unified Model for Massive NLP Tasks (NeurIPS 2023 · similarity 0.85)
- Multi-Matching Network for Multiple Choice Reading Comprehension (AAAI 2019 · similarity 0.85)
- Semantic Sentence Matching with Densely-Connected Recurrent and Co-Attentive Information (AAAI 2019 · similarity 0.85)
- Machine Comprehension Using Match-LSTM and Answer Pointer (ICLR 2017 · 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.