BID-LoRA uses bi-directional low-rank adapters with retain/new/unlearn pathways and escape unlearning to enable continual learning and unlearning while minimizing knowledge leakage and parameter updates.
Overcoming catastrophic forgetting in neural networks
8 Pith papers cite this work. Polarity classification is still indexing.
8
Pith papers citing it
citation-role summary
baseline 1
citation-polarity summary
verdicts
UNVERDICTED 8roles
baseline 1polarities
baseline 1representative citing papers
ADR achieves theoretically zero-forgetting class-incremental graph learning by combining backpropagation adaptation with ridge-regression-based layer-wise merging of GNN linear transformations.
AAND is a two-stage anomaly detection method that advances a pre-trained teacher via residual anomaly amplification and applies hard knowledge distillation in reverse distillation to achieve SOTA results on MVTecAD, VisA, and MVTec3D-RGB.
iGSP uses implicit gradient subspace projection in two phases to enable efficient continual adaptation of vision-language models, claiming SOTA accuracy with 42.7% fewer trainable parameters and 86.9% less total parameter growth.
Proposes a domain incremental continual learning benchmark for ICU time series model transportability across US regions and evaluates data replay and EWC methods.
AFU-IC decouples client unlearning from global federated training in medical imaging and adds server-side invariance calibration to prevent relearning of erased data.
RETROFIT enables continual learning for malware detection and binary summarization by retrospective-free parameter merging with low-rank sparse updates and confidence-guided arbitration, improving retention and generalization without historical data.
UW-ER integrates predictive uncertainty into experience replay for stable online CSI prediction in MIMO systems, showing NMSE near 0 dB and strong uncertainty-error correlation on 3GPP data.
citing papers explorer
-
BID-LoRA: A Parameter-Efficient Framework for Continual Learning and Unlearning
BID-LoRA uses bi-directional low-rank adapters with retain/new/unlearn pathways and escape unlearning to enable continual learning and unlearning while minimizing knowledge leakage and parameter updates.
-
Analytic Drift Resister for Non-Exemplar Continual Graph Learning
ADR achieves theoretically zero-forgetting class-incremental graph learning by combining backpropagation adaptation with ridge-regression-based layer-wise merging of GNN linear transformations.
-
Advancing Pre-trained Teacher: Towards Robust Feature Discrepancy for Anomaly Detection
AAND is a two-stage anomaly detection method that advances a pre-trained teacher via residual anomaly amplification and applies hard knowledge distillation in reverse distillation to achieve SOTA results on MVTecAD, VisA, and MVTec3D-RGB.
-
iGSP:Implicit Gradient Subspace Projection for Efficient Continual Learning of Vision-Language Models
iGSP uses implicit gradient subspace projection in two phases to enable efficient continual adaptation of vision-language models, claiming SOTA accuracy with 42.7% fewer trainable parameters and 86.9% less total parameter growth.
-
A Domain Incremental Continual Learning Benchmark for ICU Time Series Model Transportability
Proposes a domain incremental continual learning benchmark for ICU time series model transportability across US regions and evaluates data replay and EWC methods.
-
Asynchronous Federated Unlearning with Invariance Calibration for Medical Imaging
AFU-IC decouples client unlearning from global federated training in medical imaging and adds server-side invariance calibration to prevent relearning of erased data.
-
Retrofit: Continual Learning with Controlled Forgetting for Binary Security Detection and Analysis
RETROFIT enables continual learning for malware detection and binary summarization by retrospective-free parameter merging with low-rank sparse updates and confidence-guided arbitration, improving retention and generalization without historical data.
-
Uncertainty-Weighted Experience Replay for Continual MIMO Channel Prediction
UW-ER integrates predictive uncertainty into experience replay for stable online CSI prediction in MIMO systems, showing NMSE near 0 dB and strong uncertainty-error correlation on 3GPP data.