Research map / Large language models / Retrieval and knowledge
Faithful Embeddings for Knowledge Base Queries
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 EmQL instead of Query2Box | KB query embedding · FB15k entailment | Hits@3 (Avg): 43.7 -> 91.4 | Helped |
| Use EmQL instead of Query2Box | KB query embedding · FB15k-237 entailment | Hits@3 (Avg): 36.2 -> 94.2 | Helped |
| Use EmQL instead of Query2Box | KB query embedding · NELL995 entailment | Hits@3 (Avg): 51.1 -> 98.8 | Helped |
| Use EmQL instead of Query2Box | KB query embedding · FB15k generalization | Hits@3 (Avg): 48.4 -> 49.5 | Mixed |
| Use EmQL instead of Query2Box | KB query embedding · FB15k-237 generalization | Hits@3 (Avg): 26.8 -> 35.8 | Helped |
| Use EmQL instead of Query2Box | KB query embedding · NELL995 generalization | Hits@3 (Avg): 30.6 -> 46.8 | Helped |
| Add count-min sketches to the set representation (ablation) | EmQL · FB15k entailment | Hits@3 (Avg): 55.5 -> 91.4 | Helped |
| Add count-min sketches to the set representation (ablation) | EmQL · FB15k-237 entailment | Hits@3 (Avg): 57.4 -> 94.2 | Helped |
| Add count-min sketches to the set representation (ablation) | EmQL · NELL995 entailment | Hits@3 (Avg): 82.5 -> 98.8 | Helped |
| Add count-min sketches throughout the QA model (ablation) | EmQL on WebQuestionsSP · WebQuestionsSP | Hits@1: 53.2 -> 75.5 | Helped |
| Add relational filtering (ablation) | EmQL on WebQuestionsSP · WebQuestionsSP | Hits@1: 65.2 -> 75.5 | Helped |
| Use exact rather than approximate MIPS (ablation) | EmQL on WebQuestionsSP · WebQuestionsSP | Hits@1: 73.4 -> 75.5 | Helped |
Closest papers by meaning
- Query2box: Reasoning over Knowledge Graphs in Vector Space Using Box Embeddings (ICLR 2020 · similarity 0.90)
- Neural-Symbolic Entangled Framework for Complex Query Answering (NeurIPS 2022 · similarity 0.90)
- Beta Embeddings for Multi-Hop Logical Reasoning in Knowledge Graphs (NeurIPS 2020 · similarity 0.90)
- An Evaluation of Approaches to Train Embeddings for Logical Inference (Student Abstract) (AAAI 2025 · similarity 0.89)
- Neural Methods for Logical Reasoning over Knowledge Graphs (ICLR 2022 · similarity 0.89)
- Rethinking Complex Queries on Knowledge Graphs with Neural Link Predictors (ICLR 2024 · similarity 0.89)
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.