Research map / Architectures and efficiency / Attention and transformers
Temporal Attention-Gated Model for Robust Sequence Classification
Check 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 the attention-gated recurrent representation with a feed-forward classifier over attention-weighted features (ablation) | TAGM · Arabic spoken digit, noisy | accuracy (%): 85.59 -> 97.64 | Hurt |
| Use bidirectional recurrent attention-gated units (ablation) | TAGM · Arabic spoken digit, noisy | accuracy (%): 97.64 -> 97.91 | Helped |
| Use TAGM instead of LSTM | sequence classification · SST sentence-level labels, binary | accuracy (%): 85.9 -> 86.2 | Helped |
| Use TAGM instead of LSTM | sequence classification · SST sentence-level labels, fine-grained | accuracy (%): 47.2 -> 48.0 | Helped |
| Use TAGM instead of the strongest listed baseline | sequence classification · SST phrase- and sentence-level labels, binary | accuracy (%): 88.1 -> 87.6 | Hurt |
| Use TAGM instead of the strongest listed baseline | sequence classification · SST phrase- and sentence-level labels, fine-grained | accuracy (%): 50.6 -> 50.1 | Hurt |
| Use TAGM instead of TreeLSTM | sequence classification · SST phrase- and sentence-level labels | overall performance: 137.5 -> 137.7 | Helped |
| Use TAGM instead of LSTM | video event classification · CCV | mAP: 0.55 -> 0.63 | Helped |
| Use TAGM instead of the CNN-feature BOW+SVM baseline | video event classification · CCV | mAP: 0.67 -> 0.63 | Hurt |
Closest papers by meaning
- Not All Attention Is Needed: Gated Attention Network for Sequence Data (AAAI 2020 · similarity 0.89)
- Posterior Attention Models for Sequence to Sequence Learning (ICLR 2019 · similarity 0.87)
- Mega: Moving Average Equipped Gated Attention (ICLR 2023 · similarity 0.87)
- Asymptotics of SGD in Sequence-Single Index Models and Single-Layer Attention Networks (NeurIPS 2025 · similarity 0.86)
- Explaining Modern Gated-Linear RNNs via a Unified Implicit Attention Formulation (ICLR 2025 · similarity 0.85)
- Fundamental limits of learning in sequence multi-index models and deep attention networks: high-dimensional asymptotics and sharp thresholds (ICML 2025 · 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.