Research map / Architectures and efficiency / Attention and transformers
Are Transformers universal approximators of sequence-to-sequence functions?
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 |
|---|---|---|---|
| Replace the first 1 or 2 self-attention layers with depth-wise separable convolutions (ablation) | 12-layer BERT-based Transformer · SQuAD | performance | Helped |
| Replace the first 1 or 2 self-attention layers with depth-wise separable convolutions (ablation) | 12-layer BERT-based Transformer · MNLI | performance | Helped |
| Replace more self-attention layers with depth-wise separable convolutions (ablation) | 12-layer BERT-based Transformer · SQuAD and MNLI | performance | Hurt |
| Use bilinear projection or depth-wise separable convolution instead of self-attention throughout (ablation) | Standalone alternative-layer models · MNLI | performance | Hurt |
Closest papers by meaning
- Prompting a Pretrained Transformer Can Be a Universal Approximator (ICML 2024 · similarity 0.91)
- Understanding the Expressive Power and Mechanisms of Transformer for Sequence Modeling (NeurIPS 2024 · similarity 0.90)
- Approximation Rate of the Transformer Architecture for Sequence Modeling (NeurIPS 2024 · similarity 0.88)
- Transformers are Universal In-context Learners (ICLR 2025 · similarity 0.88)
- The Power of Hard Attention Transformers on Data Sequences: A formal language theoretic perspective (NeurIPS 2024 · similarity 0.88)
- Are Transformers with One Layer Self-Attention Using Low-Rank Weight Matrices Universal Approximators? (ICLR 2024 · similarity 0.88)
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.