Fusing Pre-Trained Language Models with Multimodal Prompts through Reinforcement Learning.
Youngjae YuJiwan ChungHeeseung YunJack HesselJae Sung ParkXiming LuRowan ZellersPrithviraj AmmanabroluRonan Le BrasGunhee KimYejin ChoiPublished in: CVPR (2023)
Keyphrases
- language model
- pre trained
- reinforcement learning
- language modeling
- training data
- probabilistic model
- n gram
- information retrieval
- document retrieval
- speech recognition
- retrieval model
- test collection
- smoothing methods
- query expansion
- training examples
- statistical language models
- language modelling
- control signals
- machine learning
- multi modal
- state space
- document ranking
- language models for information retrieval
- relevance model
- prior knowledge
- learning algorithm
- data sets
- supervised learning
- active learning
- fusion method
- learning process
- multimedia