INVITE: a Testbed of Automatically Generated Invalid Questions to Evaluate Large Language Models for Hallucinations.
Anil RamakrishnaRahul GuptaJens LehmannMorteza ZiyadiPublished in: EMNLP (Findings) (2023)
Keyphrases
- automatically generated
- language model
- language modeling
- automatically generate
- probabilistic model
- manually created
- speech recognition
- information retrieval
- context sensitive
- retrieval model
- n gram
- language modelling
- query expansion
- manually generated
- document retrieval
- test collection
- smoothing methods
- manually constructed
- statistical language models
- query terms
- ad hoc information retrieval
- automatically created
- vector space model
- document ranking
- translation model
- language model for information retrieval
- open directory project
- relevance model
- okapi bm
- term dependencies
- multiword
- feature selection