NeurIPS 2026 accepted papers on Continual learning
59 of the 6,080 papers accepted at NeurIPS 2026 (1.0%) are on Continual learning, compared with 0.9% at NeurIPS 2025. They fall into 4 topics.
See NeurIPS 2026 in the interactive map
Working on something in this area? Describe your idea in scime atlas to see the closest accepted papers by meaning and how crowded the spot has become.
Topics and papers
Continual fine-tuning of LLMs · 24 papers
- Dynamic Cross-Modal Prompt Generation for Multimodal Continual Instruction Tuning
- Estimating and Orthogonalizing Unknown Pre-training Gradients for Continual Fine-tuning of Large Language Models
- FLAME: Adaptive Mixture-of-Experts for Continual Multimodal Multi-Task Learning
- Geometry Conflict: Explaining and Controlling Forgetting in LLM Continual Post-Training
- InduceKV: Fixed-Footprint Continual Adaptation of Multimodal LLMs via Inducing KV Memories
- Overcoming Catastrophic Forgetting in Visual Continual Learning with Reinforcement Fine-Tuning
- PaLoRA: Paced Low-Rank Adaptation for Continual Learning
- Rosetta: Composable Native Multimodal Pretraining
- Tell Me What To Learn: Generalizing Neural Memory to be Controllable in Natural Language
- When the Merge Coefficient Stops Mattering: Proximity Regularized Merging for Continual LoRA Adaptation
- AdaSRU: Adaptive Source-Free Recommendation Unlearning via Gradient-Constrained Optimization
- C-LoRA: Continual Low-Rank Adaptation for Pre-trained Visual Models
and 12 more in the interactive map.
Online and replay-based · 19 papers
- Autonomous Continual Learning for Environment Adaptation of Computer-Use Agents
- Continual Learning Bench: Evaluating Frontier AI Systems in Real-World Stateful Environments
- Continual Learning in Modern Hopfield Networks with an Application to Diffusion Models
- Forgetting Has Neighbors: Localized Collateral Forgetting in Machine Unlearning
- InTAct: Interval-based Task Activation Consolidation for Continual Learning
- Learning to Continually Learn via Meta-learning Agentic Memory Designs
- MoRe: Modular Representations for Principled Continual Representation Learning on Sequential Data
- Online Learning via Learned Latent Bayesian Tracking
- Re-evaluating Continual Learning with Few-Shot Adaptation
- Trust Region Continual Learning as an Implicit Meta-Learner
- ALIGN-Rec: Continual Recommendation under Heterogeneous Unlearning Requests
- Bayes-pFCL:Bayesian Personalized Federated Continual Learning
and 7 more in the interactive map.
Class-incremental · 11 papers
- Data-Free Reservoir Features for Efficient Long-Horizon Cold-Start Continual Learning
- Dynamical Adapter Fusion: Constructing A Global Adapter for Pre-Trained Model-based Class-Incremental Learning
- Hierarchical Semantic Tree Anchoring for CLIP-Based Class-Incremental Learning
- MIST: Reliable Streaming Decision Trees for Online Class-Incremental Learning via McDiarmid Bound
- A Theoretical Bridge Between Long-Tailed Recognition and Continual Learning
- Beyond Exemplar Selection: Value-Aware Memory Allocation in Replay-Based Continual Learning
- Class-Domain Incremental Learning with Extensible Multi-Center Modeling
- Gram-Calibrated Anchoring for Class-Incremental Learning
- Hyperbolic Displacement-Constrained Adaptation for CLIP-Based Class-Incremental Learning
- Subspace-Guided Continual Learning: Hessian Based Stable–Plastic Decomposition for Exemplar-Free Class-Incremental Learning
- TailAdapt: Heavy-Tailed Sparse Variational Adaptation for Long-Tailed Class Incremental Learning
Domain and prompt-based · 5 papers
- Anchoring Physiological Invariance, Expanding Domain Plasticity: Prior-Stabilized Dynamic Adaptation for Continual rPPG Measurement
- Cross-Domain Knowledge Separation and Positive Transmission for Noisy Domain Incremental Learning
- CTRL: Continual Test-Time Reinforcement Learning for Large Language Models
- Reliability-Budgeted Edge–Cloud Adaptation for Continual Multimodal Dehazing on UAV
- Towards Better Generalization in Lifelong Person Re-Identification with Flatness-Aware Learning