Research map / Vision-language and multimodal / Visual question answering
One Question Answering Model for Many Languages with Cross-lingual Dense Passage Retrieval
Akari Asai, Xinyan Yu, Jungo Kasai, Hannaneh Hajishirzi
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 |
|---|---|---|---|
| Add iterative training and data mining to the initial annotated-data training (ablation) | CORA · XOR-TyDi QA development | Avg. F1: 27.9 -> 31.4 | Improved |
| Add iterative training and data mining to the initial annotated-data training (ablation) | CORA · MKQA subset | Avg. F1: 17.3 -> 22.3 | Improved |
| Train mDPR with cross-lingual data instead of English-only Natural Questions data (ablation) | CORA · XOR-TyDi QA development | Avg. F1: 24.3 -> 31.4 | Improved |
| Train mDPR with cross-lingual data instead of English-only Natural Questions data (ablation) | CORA · MKQA subset | Avg. F1: 17.9 -> 22.3 | Improved |
| Retrieve from multilingual documents rather than English documents only (ablation) | CORA · XOR-TyDi QA development | Avg. F1: 19.1 -> 31.4 | Improved |
| Retrieve from multilingual documents rather than English documents only (ablation) | CORA · MKQA subset | Avg. F1: 20.5 -> 22.3 | Improved |
| Use multilingual answer generation instead of an extractive reader followed by answer translation (ablation) | CORA · XOR-TyDi QA development | Avg. F1: 11.2 -> 31.4 | Improved |
| Use multilingual answer generation instead of an extractive reader followed by answer translation (ablation) | CORA · MKQA subset | Avg. F1: 12.2 -> 22.3 | Improved |
| Include Spanish, Swedish, and Indonesian document embeddings (ablation) | CORA · Danish, Portuguese, and Malay questions | F1: 1.0 in Danish, 0.6 in Portuguese, and 3.4 F1 points in Malay | Improved |
Closest papers by meaning
- Cooperative Retrieval-Augmented Generation for Question Answering: Mutual Information Exchange and Ranking by Contrasting Layers (NeurIPS 2025 · similarity 0.87)
- RepLiQA: A Question-Answering Dataset for Benchmarking LLMs on Unseen Reference Content (NeurIPS 2024 · similarity 0.87)
- Rethinking Generative Large Language Model Evaluation for Semantic Comprehension (ICML 2024 · similarity 0.86)
- Chain-of-Retrieval Augmented Generation (NeurIPS 2025 · similarity 0.86)
- SuRe: Summarizing Retrievals using Answer Candidates for Open-domain QA of LLMs (ICLR 2024 · similarity 0.86)
- Making Retrieval-Augmented Language Models Robust to Irrelevant Context (ICLR 2024 · similarity 0.86)
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.
