Research map / Large language models / Retrieval and knowledge
Differentiable Reasoning on Large Knowledge Bases and Natural Language
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 NaNTP with dynamically pruned proof paths instead of enumerating all paths | Differentiable proving · WN18 training | training speed: more than two orders of magnitude | Helped |
| Add attention over known predicates for rule-predicate embeddings (ablation) | NaNTP · Kinship | MRR: 0.719 -> 0.759 | Helped |
| Add attention over known predicates for rule-predicate embeddings (ablation) | NaNTP · UMLS | MRR: 0.841 -> 0.857 | Helped |
| Add attention over known predicates for rule-predicate embeddings (ablation) | NaNTP · Nations | MRR: 0.658 -> 0.645 | Hurt |
| Add attention over known predicates for rule-predicate embeddings (ablation) | NaNTP · Countries S1 | AUC-PR: 99.98 ± 0.05 -> 100.0 ± 0.0 | No clear effect |
| Encode textual mentions rather than simply adding them as facts | NaNTP · Countries S1-S3 with generated mentions | ranking accuracy | Helped |
| Use attention over predicates for rule induction | NaNTP · Rule learning | parameter efficiency | Helped |
Closest papers by meaning
- Improving Soft Unification with Knowledge Graph Embedding Methods (ICML 2025 · similarity 0.90)
- Boule or Baguette? A Study on Task Topology, Length Generalization, and the Benefit of Reasoning Trace (NeurIPS 2026 · similarity 0.88)
- Systematic Relational Reasoning With Epistemic Graph Neural Networks (ICLR 2025 · similarity 0.87)
- Rethinking Complex Queries on Knowledge Graphs with Neural Link Predictors (ICLR 2024 · similarity 0.87)
- Graph-constrained Reasoning: Faithful Reasoning on Knowledge Graphs with Large Language Models (ICML 2025 · similarity 0.87)
- Reasoning of Large Language Models over Knowledge Graphs with Super-Relations (ICLR 2025 · similarity 0.87)
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.