Hard Prompts Made Interpretable: Sparse Entropy Regularization for Prompt Tuning with RL.
Yunseon ChoiSangmin BaeSeonghyun BanMinchan JeongChuheng ZhangLei SongLi ZhaoJiang BianKee-Eung KimPublished in: CoRR (2024)
Keyphrases
- sparsity regularization
- reinforcement learning
- elastic net
- mixed norm
- structured sparsity
- sparse approximation
- information theory
- mutual information
- information theoretic
- rank minimization
- regularization parameter
- high dimensional
- sparse representation
- sparse coding
- sparse data
- kernel matrices
- multi agent
- reinforcement learning algorithms
- information entropy
- autonomous learning
- classification rules
- markov decision processes
- linear combination
- state space
- learning process
- feature selection
- function approximation
- regularization methods
- regularization method
- temporal difference
- control group
- machine learning
- model selection
- learning algorithm