catastrophic forgetting
This term describes the phenomenon where a neural network forgets previously learned information upon acquiring new knowledge. In the context of continual learning, researchers work on strategies to mitigate this issue, ensuring that AI systems retain old knowledge while learning new tasks.
- AnaCP: Toward Upper-Bound Continual Learning via Analytic Contrastive Projection
- Auto-Compressing Networks
- Auto-Compressing Networks
- Bisecle: Binding and Separation in Continual Learning for Video Language Understanding
- Buffer layers for Test-Time Adaptation
- C-NAV: Towards Self-Evolving Continual Object Navigation in Open World
- Class-aware Domain Knowledge Fusion and Fission for Continual Test-Time Adaptation
- Compact Memory for Continual Logistic Regression
- Continual Knowledge Adaptation for Reinforcement Learning
- Continuous Subspace Optimization for Continual Learning
- Contrastive Consolidation of Top-Down Modulations Achieves Sparsely Supervised Continual Learning
- DOTA: Distributional Test-time Adaptation of Vision-Language Models
- Data Efficient Adaptation in Large Language Models via Continuous Low-Rank Fine-Tuning
- Decentralized Dynamic Cooperation of Personalized Models for Federated Continual Learning
- DevFD : Developmental Face Forgery Detection by Learning Shared and Orthogonal LoRA Subspaces
- Dual-Space Semantic Synergy Distillation for Continual Learning of Unlabeled Streams
- Dynamic Siamese Expansion Framework for Improving Robustness in Online Continual Learning
- Federated Continual Learning via Orchestrating Multi-Scale Expertise
- GeoAda: Efficiently Finetune Geometric Diffusion Models with Equivariant Adapters
- GraphKeeper: Graph Domain-Incremental Learning via Knowledge Disentanglement and Preservation
- HMVLM:Human Motion-Vision-Language Model via MoE LoRA
- Hippocampal-like Sequential Editing for Continual Knowledge Updates in Large Language Models
- Hybrid Re-matching for Continual Learning with Parameter-Efficient Tuning
- Investigating and Mitigating Catastrophic Forgetting in Medical Knowledge Injection through Internal Knowledge Augmentation Learning
- Knowledge Graph Enhanced Generative Multi-modal Models for Class-Incremental Learning
- Learn and Ensemble Bridge Adapters for Multi-domain Task Incremental Learning
- Learning Expandable and Adaptable Representations for Continual Learning
- LoRA-EnVar: Parameter-Efficient Hybrid Ensemble Variational Assimilation for Weather Forecasting
- Looking Beyond the Known: Towards a Data Discovery Guided Open-World Object Detection
- MINGLE: Mixture of Null-Space Gated Low-Rank Experts for Test-Time Continual Model Merging
- Memory Decoder: A Pretrained, Plug-and-Play Memory for Large Language Models
- Memory-Integrated Reconfigurable Adapters: A Unified Framework for Settings with Multiple Tasks
- Mitigating Forgetting in LLM Fine-Tuning via Low-Perplexity Token Learning
- Mitigating Intra- and Inter-modal Forgetting in Continual Learning of Unified Multimodal Models
- Pay Attention to Small Weights
- ReservoirTTA: Prolonged Test-time Adaptation for Evolving and Recurring Domains
- Robot-R1: Reinforcement Learning for Enhanced Embodied Reasoning in Robotics
- SPFL: Sequential updates with Parallel aggregation for Enhanced Federated Learning under Category and Domain Shifts
- SPICED: A Synaptic Homeostasis-Inspired Framework for Unsupervised Continual EEG Decoding
- STRAP: Spatio-Temporal Pattern Retrieval for Out-of-Distribution Generalization
- Self-Evolving Pseudo-Rehearsal for Catastrophic Forgetting with Task Similarity in LLMs
- Separating the 'what' and 'how' of compositional computation to enable reuse and continual learning
- Temporal-Difference Variational Continual Learning
- Train with Perturbation, Infer after Merging: A Two-Stage Framework for Continual Learning
- Turning the Tables: Enabling Backward Transfer via Causal-Aware LoRA in Continual Learning
- UniEdit: A Unified Knowledge Editing Benchmark for Large Language Models