RepLiQA: A Question-Answering Dataset for Benchmarking LLMs on Unseen Reference Content.
João MonteiroPierre-André NoëlÉtienne MarcotteSai RajeswarValentina ZantedeschiDavid VázquezNicolas ChapadosChristopher PalPerouz TaslakianPublished in: CoRR (2024)
Keyphrases
- question answering
- natural language processing
- information retrieval
- relation extraction
- syntactic information
- named entities
- information extraction
- qa clef
- multimedia
- natural language
- natural language questions
- sentence retrieval
- passage retrieval
- metadata
- question classification
- cross language
- question answering systems
- answering questions
- training set
- answer validation
- open domain question answering
- speech transcripts
- question answer pairs