paper
Unsupervised Continual Clustering via Forward-Backward Knowledge Distillation
paperactiveprovisional
unsupervised-continual-clustering-via-forward-backward-knowledge-distillation-456db579·1 events·first seen 9d agoAliases: Unsupervised Continual Clustering via Forward-Backward Knowledge Distillation, Forward-Backward Knowledge Distillation for Continual Clustering
More like this (12)
knowledge distillationSparse Subspace-to-Expert Sharing for Task-Agnostic Continual LearningDense Supervision, Sparse Updates: On the Sparsity and Geometry of On-Policy Distillationon-policy distillationIn-context Vector DistillationFlashbackCL: Mitigating Temporal Forgetting in Federated LearningLearning from the Self-future: On-policy Self-distillation for dLLMsDetecting Knowledge Gaps from Conversational AI Interactions Using Curriculum Prerequisite GraphsRethinking Dataset Distillation for Classification: Do Distilled Sets Outperform Coresets?FFR: Forward-Forward Learning for RegressionAlternating Token-Weighted UnlearningSelf-Supervised Learning
Recent events (1)
FBCC: Forward-Backward Knowledge Distillation for Unsupervised Continual Clustering
A new arXiv preprint introduces Unsupervised Continual Clustering (UCC) as a problem formulation and proposes FBCC, a method using a continual teacher network with task-specific student networks to learn sequential clustering tasks without labels or stored past data. The approach uses a dual-phase forward-backward distillation process to preserve previously discovered cluster structure while learning new ones. Experiments on four benchmark datasets show FBCC outperforms continual learning baselines in clustering accuracy while reducing catastrophic forgetting.