Inference-Time Policy Adapters (IPA): Tailoring Extreme-Scale LMs without Fine-tuning.
Ximing LuFaeze BrahmanPeter WestJaehun JungKhyathi ChanduAbhilasha RavichanderPrithviraj AmmanabroluLiwei JiangSahana RamnathNouha DziriJillian FisherBill LinSkyler HallinanLianhui QinXiang RenSean WelleckYejin ChoiPublished in: EMNLP (2023)
Keyphrases
- fine tuning
- fine tune
- viable alternative
- learning management systems
- e learning
- input output
- fine tuned
- scale space
- optimal policy
- belief networks
- asymptotically optimal
- policy makers
- multiscale
- general purpose
- probabilistic inference
- random fields
- bayesian inference
- decision theoretic
- probability distribution
- inference process
- bayesian networks
- decision making