Simple synthetic data reduces sycophancy in large language models.
Jerry W. WeiDa HuangYifeng LuDenny ZhouQuoc V. LePublished in: CoRR (2023)
Keyphrases
- synthetic data
- language model
- language modeling
- n gram
- probabilistic model
- query expansion
- information retrieval
- speech recognition
- document retrieval
- real image data
- language modelling
- retrieval model
- statistical language models
- test collection
- real world
- vector space model
- active learning
- pseudo relevance feedback
- data sets
- ad hoc information retrieval
- knn
- query terms
- context sensitive
- document ranking
- recommender systems
- bayesian networks
- language model for information retrieval