Public articles linked to the same research event.
arXiv The work proposes LINEUP: it learns a bank of reusable low-rank personalization factors, composes them through user-conditioned recall and query-dependent calibration, and restricts target-user adaptation to a tiny user code over a shared correction space (each target user optimizes only eight scalars, versus 4.19 million per-user parameters in the evaluated private-LoRA configuration); across six tasks spanning personalized classification, prediction, and generation, LINEUP leads on all 12 metrics, each averaged over three independent runs (e.g., reducing LaMP-3 RMSE by 11.4% relative to the strongest baseline), and it maintains advantages under limited history.
The work proposes LINEUP: it learns a bank of reusable low-rank personalization factors, composes them through user-conditioned recall and query-dependent calibration, and restricts target-user adaptation to a tiny user code over a shared correction space (each target user optimizes only eight scalars, versus 4.19 million per-user parameters in the evaluated private-LoRA configuration); across six tasks spanning personalized classification, prediction, and generation, LINEUP leads on all 12 metrics, each averaged over three independent runs (e.g., reducing LaMP-3 RMSE by 11.4% relative to the strongest baseline), and it maintains advantages under limited history.
The work proposes LINEUP: it learns a bank of reusable low-rank personalization factors, composes them through user-conditioned recall and query-dependent calibration, and restricts target-user adaptation to a tiny user code over a shared correction space (each target user optimizes only eight scalars, versus 4.19 million per-user parameters in the evaluated private-LoRA configuration); across six tasks spanning personalized classification, prediction, and generation, LINEUP leads on all 12 metrics, each averaged over three independent runs (e.g., reducing LaMP-3 RMSE by 11.4% relative to the strongest baseline), and it maintains advantages under limited history.
The work proposes LINEUP: it learns a bank of reusable low-rank personalization factors, composes them through user-conditioned recall and query-dependent calibration, and restricts target-user adaptation to a tiny user code over a shared correction space (each target user optimizes only eight scalars, versus 4.19 million per-user parameters in the evaluated private-LoRA configuration); across six tasks spanning personalized classification, prediction, and generation, LINEUP leads on all 12 metrics, each averaged over three independent runs (e.g., reducing LaMP-3 RMSE by 11.4% relative to the strongest baseline), and it maintains advantages under limited history.