Research map / Architectures and efficiency / RNNs and normalization
Learn to Explain Efficiently via Neural Logic Inductive 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 |
|---|---|---|---|
| Use NLIL instead of NeuralLP | KB completion · FB15K-237 | MRR: 0.24 -> 0.25 | Helped |
| Use NLIL instead of NeuralLP | KB completion · FB15K-237 | Time (mins): 250 -> 82 | Helped |
| Use NLIL instead of NeuralLP | KB completion · WN18 | MRR: 0.94 -> 0.95 | Helped |
| Use NLIL instead of NeuralLP | KB completion · WN18 | Time (mins): 54 -> 12 | Helped |
| Use NLIL instead of RotatE | KB completion · FB15K-237 | MRR: 0.34 -> 0.25 | Hurt |
| Use NLIL instead of RotatE | KB completion · WN18 | Hits@10: 95.5 -> 94.6 | Hurt |
| Learn rules with NLIL rather than use Freq | Visual Genome object classification · Visual Genome | R@1: 0.40 -> 0.51 | Helped |
| Learn rules with NLIL rather than use Freq | Visual Genome object classification · Visual Genome | R@5: 0.44 -> 0.52 | Helped |
| Use NLIL rather than MLP+RCNN | Visual Genome object classification · Visual Genome | R@1: 0.53 -> 0.51 | Hurt |
| Use NLIL rather than MLP+RCNN | Visual Genome object classification · Visual Genome | R@5: 0.81 -> 0.52 | Hurt |
| Use NLIL rather than TransE | KB completion · FB15K-237 | MRR: 0.28 -> 0.25 | Hurt |
| Use NLIL rather than TransE | KB completion · WN18 | MRR: 0.57 -> 0.95 | Helped |
Closest papers by meaning
- Neural Logic Machines (ICLR 2019 · similarity 0.89)
- Learning Logic Programs by Discovering Where Not to Search (AAAI 2023 · similarity 0.89)
- Generalisation through Negation and Predicate Invention (AAAI 2024 · similarity 0.88)
- Differentiable Rule Induction from Raw Sequence Inputs (ICLR 2025 · similarity 0.88)
- Inductive Learning of Logical Theories with LLMs: A Expressivity-graded Analysis (AAAI 2025 · similarity 0.87)
- Learning MDL Logic Programs from Noisy Data (AAAI 2024 · similarity 0.84)
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.