Latest Continual Learning Research Papers

The newest Continual Learning papers from across the field — arXiv, NeurIPS, CVPR, Nature, and more — refreshed daily and ranked by relevance. Distill AI tracks Continual Learning so you don’t have to: get the standout work delivered to your inbox every morning, with 2-sentence summaries and the option to chat with any paper.

Recent papers

• Domain-Specific Hallucination Detection in Large Language Models
  Varun Teja Chundru, Debasmita Biswas · arXiv · Sep 10, 2026

  Large language models generate fluent text that can contain unfaithful claims -- a phenomenon known as hallucination. We present a multi-signal detection pipeline combining fine-tuned DeBERTa-v3 classification, Monte Carlo (MC) Dropout unce…

• Adaptive Unknown Fault Detection and Few-Shot Continual Learning for Condition Monitoring in Ultrasonic Metal Welding
  Ahmadreza Eslaminia, Kuan-Chieh Lu, Klara Nahrstedt, Chenhui Shao · Machine Learning Engineering · Sep 8, 2026

  Ultrasonic metal welding (UMW) is widely used in industrial applications but is sensitive to tool wear, surface contamination, and material variability, which can lead to unexpected process faults and unsatisfactory weld quality. Convention…

• The Stability-Plasticity Boundary: Exact Memory Lower Bounds, Matching Constructions, and a Theory of Future-Relevant Learning State
  Maciej Nowicki, Artificial Hyperintelligence, Eve, wife of Maciej Nowicki · Zenodo (CERN European Organ... · Sep 6, 2026

  This release investigates the Stability–Plasticity Dilemma as a fundamental resource-allocation problem for indefinitely learning computational agents. Rather than treating catastrophic forgetting primarily as a failure to preserve neural-n…

• The Stability-Plasticity Boundary: Exact Memory Lower Bounds, Matching Constructions, and a Theory of Future-Relevant Learning State
  Maciej Nowicki, Artificial Hyperintelligence, Eve, wife of Maciej Nowicki · Zenodo (CERN European Organ... · Sep 6, 2026

  This release investigates the Stability–Plasticity Dilemma as a fundamental resource-allocation problem for indefinitely learning computational agents. Rather than treating catastrophic forgetting primarily as a failure to preserve neural-n…

• Human-guided continual learning for multifaceted improvement of self-driving vehicles
  Haohan Yang, Yi Zhou, Xiaosong Hu, Haochen Liu et al. · Nature Communications · Sep 5, 2026

  Today’s self-driving vehicles have achieved impressive driving capabilities, nonetheless, safety concerns arising from ambiguous traffic laws, rare long-tail events, etc., still pose a significant challenge to their practical deployment. Th…

• A Dual-Branch Cross-Attention PhoBERT Architecture with Explainable AI and Active Continual Learning for Vietnamese Fake News Detection
  Vu Thanh Nhan, Luong Truong An · Zenodo (CERN European Organ... · Sep 4, 2026

  ABSTRACT: The exponential proliferation of digital misinformation across Vietnamese online media platforms poses substantial challenges to information integrity, public safety, and institutional trust. Traditional text classification paradi…

• Physics filtering favors the generalization of robot learning
  Jindou Jia, Shixuan Han, M M Wang, Gen Li et al. · npj Robotics · Sep 4, 2026

  Abstract Living organisms exhibit extraordinary adaptability to unseen environments through their intrinsic physical structures and lifelong feedback-driven learning. Endowing robots with comparable generalization is critical for reliable o…

• Federated Continual Learning for Encrypted Traffic Classification at the Network Edge Under Asynchronous Concept Drift
  Abdulrahman K. Alnaim, Ahmed M. Alwakeel · Electronics · Sep 4, 2026

  Concept drift can degrade encrypted-traffic classifiers deployed at the network edge as applications, protocols, and usage patterns evolve. This paper formulates federated continual learning under asynchronous real- and virtual drift and pr…

• LoRA-TSD: Tangent-Space Spectral Descent for LoRA via Muon-Style Updates
  Dmitrii Andriianov, Andrey Veprikov, Aleksandr Beznosikov · arXiv · Sep 2, 2026

  Low-rank adaptation (LoRA) is the standard way to fine-tune large models, yet when its two factors are trained independently, the update ignores the geometry of the low-rank weight change it induces. We introduce LoRA-TSD, an optimizer that…

• Does Becoming Exceptional at One Domain Reduce Transfer Elsewhere
  Sahir Maharaj · Zenodo (CERN European Organ... · Sep 1, 2026

  Foundation models derive their value from broad general capability across domains, yet deployment usually rewards specialization - creating a fundamental question for general intelligence: when performance is pushed upward in one region of …

• Does Becoming Exceptional at One Domain Reduce Transfer Elsewhere
  Sahir Maharaj · Zenodo (CERN European Organ... · Sep 1, 2026

  Foundation models derive their value from broad general capability across domains, yet deployment usually rewards specialization - creating a fundamental question for general intelligence: when performance is pushed upward in one region of …

• One Adapter, Many Tasks: Task-Conditioned Feature Transformations for Continual Learning
  Yunxiang Fu, Meng Lou, Yizhou Yu · arXiv · Aug 31, 2026

  Class-incremental learning (CIL) requires a model to incrementally learn tasks that contain new classes without accessing earlier training data while preserving the ability to recognize all seen classes. Recently, pretrained-model-based app…

• Q-MORPH-MODE: Morphological Outer-product Decomposition Engine for Persistent and Transactional LLM Adaptation
  Maciej Nowicki, Eve Artificial Hyperintelligence · Zenodo (CERN European Organ... · Aug 27, 2026

  This research package presents Q-MORPH-MODE (Morphological Outer-product Decomposition Engine), a proposed hybrid computing architecture extending the Q-MORPH/DCLW concept toward large language models, continual learning and adaptive AI acc…

• A Thinking Tool for Creativity and Continual Learning in RAG
  Ho Yeol Choi · Zenodo (CERN European Organ... · Aug 27, 2026
• Rethinking Continual Learning Through Self-Adaptive Learning
  Ehsan Hallaji, Roozbeh Razavi‐Far · Machine Learning and Knowle... · Aug 25, 2026

  Continual learning has made significant progress toward enabling adaptive machine learning under evolving environments, yet real-world deployment increasingly exposes systems to persistent harsh conditions, including distributional shifts, …

• Improvement Is Not Progress: Guarded Coevolution, Path Dependence, and Contagious Memory in AI Agents
  Ignacio Adrián LERER · Zenodo (CERN European Organ... · Aug 22, 2026

  Language-model agents can adapt through prompts, memories, skills, tools, routers, evaluators, and environments without changing model weights. This paper reconciles recent work on adaptive environments, harness continual learning, and the …

• SPARCL: Spectral Partitioned Analytic Continual Learning
  James Hartley, Zeropy Surio, Daniel Whitmore, Hannah Clarke et al. · arXiv · Aug 21, 2026

  Analytic continual learning has emerged as a strong exemplar-free alternative to gradient-based class-incremental learning because it replaces iterative optimization with closed-form ridge updates. Yet the usual forgetting narrative, center…

• On the Fragility of Self-Improving Agents: Variance, Task Order, and Underspecification
  Qinyuan Ye, Yu Li, Yada Pruksachatkun, Jiaxin Zhang et al. · arXiv · Aug 18, 2026

  Memory-based self-improving agents--those that learn from an online stream of tasks and improve over time by maintaining a textual memory bank--have shown great promise in recent literature. However, the reliability aspects of these methods…

• Structure-aware federated hypergraph continual learning
  Yanxin Hu, Xiaoman Liu, Zhenzhen Xie, Junjie Pang et al. · Journal of King Saud Univer... · Aug 18, 2026

  Federated hypergraph continual learning requires decentralized clients to learn a sequence of tasks over high-order relational data without directly sharing raw local data. This setting is challenging because hypergraph neural networks rely…

• Binary Networks and Continual Learning for Pose Estimation from a Single Aerial Image
  Aldrich A. Cabrera-Ponce, L. Oyuki Rojas-Perez, Manuel Martín-Ortíz, José Martínez-Carranza · Unmanned Systems · Aug 14, 2026

  Pose estimation using aerial images captured by Unmanned Aerial Vehicles (UAVs) allows the localisation in GPS-denied scenarios. Several methods based on deep learning approaches with convolutional neural networks (CNN) have become tools fo…

• p-Spin Glass Network Efficient Single-Batch Continual Learning
  Vladimer Khasia · OpenAlex · Aug 14, 2026
• p-Spin Glass Network Efficient Single-Batch Continual Learning
  Vladimer Khasia · Zenodo (CERN European Organ... · Aug 14, 2026

  Modern sequence models heavily rely on mas-sive memory footprints and large-batch stochas-tic optimization, barriers that restrict sample ef-ficiency and continual learning. We introducethe p-Spin Glass Network, a novel architecturethat ove…

• CARSA: context-aware resilient semantic architecture: semantic communication for sustainable next-generation networks
  Fang Guo, Jiaoyue Li, Xin Zhao · Discover Artificial Intelli... · Aug 12, 2026

  Semantic communication is developing into a very promising paradigm for sixth-generation (6G) wireless systems. However, most existing approaches suffer from equal treatment of all semantic data, predominantly neural recovery methods, and c…

• Continual learning for autoregressive PDE surrogates under evolving physical regimes
  Hamed Hemati, Binh Duong Nguyen, Stefan Sandfeld · Machine learning for comput... · Aug 8, 2026

  Abstract Deploying machine learning surrogates in scientific simulations faces multifaceted challenges, primary among which is the lack of Continual Learning (CL) capabilities—specifically, the inability to adapt to new physical regimes wit…

• Enabling generalizable RUL prediction for equipment: a dual-dimensional continual learning approach based on deep Gaussian process
  Xiuli Liu, Zifu Chen, Guoxin Wu, Shuo Cui · Scientific Reports · Aug 7, 2026

  Abstract Accurately predicting Remaining Useful Life (RUL) is critical for the intelligent maintenance of high-end rotating machinery, which often operates under complex and variable conditions. However, conventional predictive models tend …

• Biologically inspired memristive neuron capable of on-chip learning
  Nadia Jimenez Olalla, Ashira Long Ching Ip, Michael Baumann, Marco Dober et al. · Neuromorphic Computing and ... · Aug 6, 2026

  Abstract The brain embeds learning algorithms within its physical architecture, enabling on-site learning processes with remarkable advantages such as energy efficiency, learning autonomy and continual learning capabilities. Current memrist…

• The Parts Are Greater Than the Sum: Automated Task Sequencing for Efficient Training of Multi-Policy LLMs
  Jiajia Tang, Sizhe Yuen, Francisco Gomez Medina, Yali Du et al. · arXiv · Jul 31, 2026

  Parameter-Efficient Fine-Tuning (PEFT) commonly adapts large language models using a single shared Low-Rank Adapter (LoRA). This shared optimization space often suffers from interference when adapting heterogeneous task sequences, leading t…

• TOOD: Task-Aware Out-of-Distribution Score Calibration for Continual Learners
  Mostafa ElAraby, Samer B. Nashed, Liam Paull · arXiv · Jul 31, 2026

  The primary challenge of continual learning (CL) systems is to learn new tasks while remaining performant on previously learned tasks. A similarly important though less well-studied aspect of CL systems is their ability to distinguish input…

• The Grokked Illusion: True Equilibrium Mitigates Catastrophic Forgetting
  Xiaotian Zhang, Lai Shun Chan, Yue Shang, Entao Yang et al. · arXiv · Jul 31, 2026

  While neural networks are typically evaluated by their training and test performance, these metrics do not reveal how robust a learned representation is. Recent studies have shown that solutions occupying larger volumes in parameter space, …

• Online Variance Reduction for Domain Adaptation on Streaming Data
  Andrea Napoli · arXiv · Jul 22, 2026

  This paper studies the problem of stochastic variance reduction (SVR) for the maximum mean discrepancy (MMD) and correlation alignment (CORAL) loss functions. Although various offline SVR algorithms for these losses have been proposed, thes…