Research map / Vision-language and multimodal / Visual question answering
HAS-QA: Hierarchical Answer Spans Model for Open-Domain Question Answering
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 MAX aggregation over multiple answer spans instead of selecting only the first span (ablation) | HAS-QA · QuasarT | EM: 0.372 -> 0.432 | Helped |
| Use SUM aggregation over multiple answer spans instead of selecting only the first span (ablation) | HAS-QA · QuasarT | EM: 0.372 -> 0.423 | Helped |
| Randomly select an answer span for aggregation instead of using MAX (ablation) | HAS-QA · QuasarT | EM: 0.432 -> 0.341 | Hurt |
| Use SUM aggregation instead of MAX aggregation (ablation) | HAS-QA · QuasarT | EM: 0.423 -> 0.432 | No clear effect |
| Increase the number of candidate start positions in beam search (ablation) | HAS-QA · QuasarT | EM: 0.428 -> 0.432 | Helped |
| Increase the number of candidate end positions in beam search (ablation) | HAS-QA · QuasarT | EM: 0.428 -> 0.428 | No clear effect |
| Condition end-position prediction directly on the selected start position | HAS-QA · QuasarT | EM | Helped |
| Add paragraph quality probabilities to weight evidence paragraphs (ablation) | HAS-QA · QuasarT | EM | Helped |
Closest papers by meaning
- SuRe: Summarizing Retrievals using Answer Candidates for Open-domain QA of LLMs (ICLR 2024 · similarity 0.89)
- A Semantic-Sampling Framework for Evaluating Calibration in Open-Ended Question Answering (NeurIPS 2026 · similarity 0.87)
- RepLiQA: A Question-Answering Dataset for Benchmarking LLMs on Unseen Reference Content (NeurIPS 2024 · similarity 0.86)
- Rethinking Generative Large Language Model Evaluation for Semantic Comprehension (ICML 2024 · similarity 0.86)
- ChatQA: Surpassing GPT-4 on Conversational QA and RAG (NeurIPS 2024 · similarity 0.85)
- Explore What LLM Does Not Know in Complex Question Answering (AAAI 2025 · 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.