Research map / Large language models / Retrieval and knowledge
Combining Axiom Injection and Knowledge Base Completion for Efficient Natural Language Inference
Masashi Yoshikawa, Koji Mineshima, Hiroshi Noji, Daisuke Bekki
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| Replace search-based abduction with ComplEx KBC scoring | ccg2lambda with WordNet · SICK trial | Speed (sec.): 9.15 -> 4.03 | Improved |
| Replace search-based abduction with ComplEx KBC scoring | ccg2lambda with WordNet · SICK test | accuracy: 83.55 -> 83.55 | No clear effect |
| Use the abduction tactic to avoid restarting Coq | Search-based abduction with WordNet and VerbOcean · SICK trial | Speed (sec.): 9.42 -> 7.07 | Improved |
| Add VerbOcean triplets (ablation) | KBC-based abduction with WordNet · SICK test | Recall: 64.38 -> 64.47 | Improved |
| Add VerbOcean triplets (ablation) | KBC-based abduction with WordNet · SICK test | accuracy: 83.55 -> 83.45 | Hurt |
| Add VerbOcean triplets (ablation) | KBC-based abduction with WordNet · SICK test | F1: 77.14 -> 77.04 | Hurt |
| Add VerbOcean triplets (ablation) | KBC-based abduction with WordNet · SICK trial | Speed (sec.): 4.03 -> 3.84 | Improved |
| Use KBC-based rather than search-based abduction | ccg2lambda with WordNet and VerbOcean · LexSICK | Correct: 20 / 58 -> 21 / 58 | Improved |
| Use ComplEx rather than ConvE | WordNet knowledge base completion · WordNet development triplets | MRR: 67.41 -> 77.68 | Improved |
Closest papers by meaning
- LogicCat: A Chain-of-Thought Text-to-SQL Benchmark for Complex Reasoning (AAAI 2026 · similarity 0.86)
- Are Language Models Efficient Reasoners? A Perspective from Logic Programming (NeurIPS 2025 · similarity 0.85)
- Enhancing Reasoning Capabilities of LLMs via Principled Synthetic Logic Corpus (NeurIPS 2024 · similarity 0.85)
- ORACLE: Optimizing Reasoning Abilities of Large Language Models via Constraint-Led Synthetic Data Elicitation (AAAI 2026 · similarity 0.85)
- Finding the Minimal Parameter Budget for Implicit Reasoning: A Data Complexity Driven Scaling Law for Language Models (ICML 2026 · similarity 0.85)
- Efficient Reasoning for Large Reasoning Language Models via Certainty-Guided Reflection Suppression (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.
