Research map / Vision-language and multimodal / Visual question answering
Benchmarking Knowledge-Enhanced Commonsense Question Answering via Knowledge-to-Text Transformation
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 |
|---|---|---|---|
| Add ConceptNet knowledge using the full knowledge-to-text framework | BERT-Large · CommonsenseQA | accuracy: 63.6 -> 70.4 | Helped |
| Add ConceptNet knowledge using the full knowledge-to-text framework | XLNet-Large · CommonsenseQA | accuracy: 68.9 -> 80.3 | Helped |
| Add ConceptNet knowledge using the full knowledge-to-text framework | RoBERTa-Large · CommonsenseQA | accuracy: 76.2 -> 80.8 | Helped |
| Add ConceptNet knowledge using the full knowledge-to-text framework | ALBERT-XXLarge · CommonsenseQA | accuracy: 78.6 -> 83.3 | Helped |
| Combine template, paraphrasing, and retrieval descriptions (ablation) | Knowledge-to-text model · CommonsenseQA | accuracy: 67.9 -> 70.4 | Helped |
| Use template-based descriptions rather than paraphrased descriptions (ablation) | Knowledge-to-text model · CommonsenseQA | accuracy: 67.2 -> 67.9; 74.9 -> 77.5; 77.8 -> 78.1; 79.3 -> 81.1 | Helped |
| Use template-based descriptions rather than retrieved Wikipedia descriptions (ablation) | Knowledge-to-text model · CommonsenseQA | accuracy: 65.0 -> 67.9; 75.0 -> 77.5; 77.1 -> 78.1; 79.4 -> 81.1 | Helped |
| Add template-based ConceptNet knowledge | BERT · WSC | accuracy: 66.0 -> 68.1 | Helped |
| Add template-based ConceptNet knowledge | BERT · HellaSWAG dev | overall accuracy: 42.3 -> 44.2 | Helped |
| Add template-based ConceptNet knowledge | BERT · SOCIAL IQa | accuracy: 66.2 -> 68.8 | Helped |
| Add template-based ConceptNet knowledge | RoBERTa · WSC | accuracy: 81.4 -> 82.5 | Helped |
| Add template-based ConceptNet knowledge | RoBERTa · HellaSWAG dev | overall accuracy: 82.5 -> 83.0 | Helped |
Closest papers by meaning
- FlexKBQA: A Flexible LLM-Powered Framework for Few-Shot Knowledge Base Question Answering (AAAI 2024 · similarity 0.88)
- Code-Style In-Context Learning for Knowledge-Based Question Answering (AAAI 2024 · similarity 0.88)
- KBQA-o1: Agentic Knowledge Base Question Answering with Monte Carlo Tree Search (ICML 2025 · similarity 0.87)
- NovelQA: Benchmarking Question Answering on Documents Exceeding 200K Tokens (ICLR 2025 · similarity 0.87)
- SuRe: Summarizing Retrievals using Answer Candidates for Open-domain QA of LLMs (ICLR 2024 · similarity 0.87)
- TrustUQA: A Trustful Framework for Unified Structured Data Question Answering (AAAI 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.