A representation learning approach for multi-source domain adaptation achieves identifiability by partitioning the label's Markov blanket into parents, children, and spouses.
Title resolution pending
4 Pith papers cite this work. Polarity classification is still indexing.
4
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
verdicts
UNVERDICTED 4roles
background 1polarities
background 1representative citing papers
Pilot study uses pretrained video encoder features from lung ultrasound to predict 30-day CHF readmission, finding lower-lung views and temporal differences most informative with top MLP F1 of 0.80.
Causal mediation analysis shows harmful LLM outputs arise in late layers from MLP failures and gating neurons, with early layers handling harm context detection and signal propagation.
Identifiable latent bandits apply nonlinear ICA to observational data to recover representations sufficient for inferring optimal actions in new instances, shortening exploration time.
citing papers explorer
No citing papers match the current filters.