Research map / Learning theory and algorithms / Combinatorial and search optimization
Learning Logic Programs by Discovering Where Not to Search
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 |
|---|---|---|---|
| Discover BK constraints before hypothesis search | Popper on buttons · buttons | learning time: 686s to 25s; -96% | Helped |
| Discover BK constraints before hypothesis search | Popper on IGGP md · md | learning time: 142 ± 7 to 13 ± 0.4 | Helped |
| Discover BK constraints before hypothesis search | Popper on krk · krk | learning time: 48 ± 6 to 9 ± 0.6 | Helped |
| Discover BK constraints before hypothesis search | Popper on imdb3 · imdb3 | learning time: 366s to 287s; 21% reduction | Helped |
| Discover BK constraints before hypothesis search | Popper across evaluated tasks · multiple domains | learning time: up to 97% | Helped |
| Discover BK constraints before hypothesis search | Popper on sumlist · sumlist | predictive accuracy: 90 ± 6 to 100 ± 0 | Helped |
| Discover BK constraints before hypothesis search | Popper on imdb1 and imdb2 · IMDB | predictive accuracy: imdb1: 100 ± 0 to 100 ± 0; imdb2: 100 ± 0 to 100 ± 0 | No clear effect |
| Discover BK constraints before hypothesis search | Popper on md with maximum rule size 8 · md | learning time: timeout to 47 ± 2 | Helped |
| Discover BK constraints before hypothesis search | Popper on rps · rps | programs generated: 6297 ± 7 to 988 ± 2; -84% | Helped |
| Discover BK constraints from BK | Disco on synthesis background knowledge · synthesis BK | constraint discovery time: 4.00 ± 0.40; around 47s for around 8m facts | Mixed |
Closest papers by meaning
- Learning MDL Logic Programs from Noisy Data (AAAI 2024 · similarity 0.86)
- Hypothesis Search: Inductive Reasoning with Language Models (ICLR 2024 · similarity 0.86)
- Inductive Learning of Logical Theories with LLMs: A Expressivity-graded Analysis (AAAI 2025 · similarity 0.86)
- When Is Inductive Inference Possible? (NeurIPS 2024 · similarity 0.85)
- Phenomenal Yet Puzzling: Testing Inductive Reasoning Capabilities of Language Models with Hypothesis Refinement (ICLR 2024 · similarity 0.84)
- MIRAGE: Evaluating and Explaining Inductive Reasoning Process in Language Models (ICLR 2025 · 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.