Research map / Image generation / Text-to-image diffusion
Table-to-Text Generation by Structure-Aware Seq2seq Learning
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 field embedding by concatenation (ablation) | Seq2seq · WIKIBIO test | BLEU-4: 42.06 ± 0.32 -> 43.34 ± 0.37 | Helped |
| Add field embedding by concatenation (ablation) | Seq2seq · WIKIBIO test | ROUGE-4: 38.06 ± 0.36 -> 39.84 ± 0.32 | Helped |
| Add position embedding by concatenation (ablation) | Seq2seq · WIKIBIO test | BLEU-4: 42.06 ± 0.32 -> 43.65 ± 0.44 | Helped |
| Add position embedding by concatenation (ablation) | Seq2seq · WIKIBIO test | ROUGE-4: 38.06 ± 0.36 -> 40.32 ± 0.23 | Helped |
| Use field-gating encoder (ablation) | Seq2seq · WIKIBIO test | BLEU-4: 42.06 ± 0.32 -> 43.74 ± 0.23 | Helped |
| Use field-gating encoder (ablation) | Seq2seq · WIKIBIO test | ROUGE-4: 38.06 ± 0.36 -> 40.53 ± 0.31 | Helped |
| Add dual attention (ablation) | Field-gating Seq2seq · WIKIBIO test | BLEU-4: 43.74 ± 0.23 -> 44.89 ± 0.33 | Helped |
| Add dual attention (ablation) | Field-gating Seq2seq · WIKIBIO test | ROUGE-4: 40.53 ± 0.31 -> 41.21 ± 0.25 | Helped |
| Use beam search (k=5) (ablation) | Structure-aware Seq2seq + dual attention · WIKIBIO test | BLEU-4: 44.89 ± 0.33 -> 44.71 | Hurt |
| Use beam search (k=5) (ablation) | Structure-aware Seq2seq + dual attention · WIKIBIO test | ROUGE-4: 41.21 ± 0.25 -> 41.65 | Helped |
| Shuffle table records | Seq2seq · WIKIBIO test | BLEU: -2.02 | Hurt |
| Shuffle table records | Seq2seq + field & position · WIKIBIO test | BLEU: -1.55 | Hurt |
Closest papers by meaning
- TableRAG: Million-Token Table Understanding with Language Models (NeurIPS 2024 · similarity 0.84)
- NeedleInATable: Exploring Long-Context Capability of Large Language Models towards Long-Structured Tables (NeurIPS 2025 · similarity 0.84)
- HeGTa: Leveraging Heterogeneous Graph-enhanced Large Language Models for Few-shot Complex Table Understanding (AAAI 2025 · similarity 0.84)
- SynTab-LLaVA: Enhancing Multimodal Table Understanding with Decoupled Synthesis (CVPR 2025 · similarity 0.83)
- Table2LaTeX-RL: High-Fidelity LaTeX Code Generation from Table Images via Reinforced Multimodal Language Models (NeurIPS 2025 · similarity 0.83)
- Precise Information Control in Long-Form Text Generation (NeurIPS 2025 · similarity 0.82)
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.