Research map / Architectures and efficiency / Attention and transformers
Graphix-T5: Mixing Pre-trained Transformers with Graph-Aware Layers for Text-to-SQL Parsing
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 Graphix graph-aware layers to the T5 encoder | T5-large · Spider dev | EM: 67.0 -> 72.7 | Helped |
| Add Graphix graph-aware layers to the T5 encoder | T5-large · Spider dev | EX: 69.3 -> 75.9 | Helped |
| Add Graphix graph-aware layers to the T5 encoder | T5-3B · Spider dev | EM: 71.5 -> 75.6 | Helped |
| Add Graphix graph-aware layers to the T5 encoder | T5-3B · Syn | EM: 58.0 -> 66.9 | Helped |
| Add Graphix graph-aware layers to the T5 encoder | T5-3B · Dk | EM: 46.9 -> 51.2 | Helped |
| Add Graphix graph-aware layers to the T5 encoder | T5-3B · Realistic | EM: 62.0 -> 72.4 | Helped |
| Add Graphix graph-aware layers to the T5 encoder | T5-3B · Spider-SSP Template | EM: 64.8 -> 70.1 | Helped |
| Add constrained decoding with PICARD | Graphix-T5-3B · Spider dev | EM: 75.6 -> 77.1 | Helped |
| Replace Bridge Mode with No-Match Mode (ablation) | Graphix-T5-large · Spider dev | EM: 72.7 -> 71.1 | Hurt |
| Replace Bridge Mode with No-Match Mode (ablation) | Graphix-T5-large · Spider dev | EX: 75.9 -> 74.2 | Hurt |
| Add Graphix layers to the decoder in addition to the encoder (ablation) | Graphix-T5-large · Spider dev | EM: 72.7 -> 72.0 | Hurt |
| Add Graphix layers to the decoder in addition to the encoder (ablation) | Graphix-T5-large · Spider dev | EX: 75.9 -> 74.7 | Hurt |
Closest papers by meaning
- GraPPa: Grammar-Augmented Pre-Training for Table Semantic Parsing (ICLR 2021 · similarity 0.88)
- CogSQL: A Cognitive Framework for Enhancing Large Language Models in Text-to-SQL Translation (AAAI 2025 · similarity 0.87)
- Alpha-SQL: Zero-Shot Text-to-SQL using Monte Carlo Tree Search (ICML 2025 · similarity 0.86)
- SQLens: An End-to-End Framework for Error Detection and Correction in Text-to-SQL (NeurIPS 2025 · similarity 0.86)
- SQLFixAgent: Towards Semantic-Accurate Text-to-SQL Parsing via Consistency-Enhanced Multi-Agent Collaboration (AAAI 2025 · similarity 0.85)
- LogicCat: A Chain-of-Thought Text-to-SQL Benchmark for Complex Reasoning (AAAI 2026 · 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.