paper
Multi-Task Bayesian In-Context Learning
paperactiveprovisional
multi-task-bayesian-in-context-learning-ba9d0172·1 events·first seen 47h agoAliases: Multi-Task Bayesian In-Context Learning
Co-occurring entities
More like this (12)
In-Context Multiple Instance LearningMulti-Task Learningin-context learningReinforcement Learning Elicits Contextual Learning of Unseen Language TranslationLatent Context Language ModelsIn-Context Reward AdaptationWhen English Isn't the Best Teacher: Source Language Effects in Cross-Lingual In-Context LearningSparse Subspace-to-Expert Sharing for Task-Agnostic Continual LearningContextual RetrievalMultimodal LearningObserve-and-Act Adaptive Context SelectionMultiple Instance Learning (MIL)
Recent events (1)
Multi-Task Bayesian In-Context Learning for Amortized Hierarchical Inference
A new arXiv preprint introduces a multi-task in-context learning framework for amortized hierarchical Bayesian predictive inference, representing prior information as a prefix of in-context datasets fed to a transformer. The model learns to adapt predictions across families of priors, addressing the brittleness of prior-data fitted models under distribution shift. On evaluations including out-of-meta-distribution priors and high-dimensional latent structures, the method matches oracle Bayesian predictors while being orders of magnitude faster, with a real-world spatiotemporal temperature prediction demonstration.