id string | sources list | title string | abstract string | authors list | categories list | fields_of_study list | published_date timestamp[s] | url string | pdf_url string | arxiv_id string | doi string | citation_count int64 | influential_citation_count int64 | has_code bool | code_url string | venue string | quality_score float64 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
7ae33abe73137c9bf3e381e7b9cd258595066c50888b5cb89bd1bc1636bee602 | [
"arxiv",
"semantic_scholar"
] | Theoretical Foundations of Continual Learning via Drift-Plus-Penalty | In many real-world settings, data streams are nonstationary and arrive sequentially, requiring learning systems to adapt continuously without retraining from scratch. Continual learning (CL) addresses this challenge by incorporating new tasks while mitigating catastrophic forgetting, where learning new information degr... | [
"Nazreen Shah",
"Govinda Arya",
"Bharath B. N.",
"Ranjitha Prasad"
] | [
"cs.LG"
] | [
"Computer Science"
] | 2026-06-07T00:00:00 | https://arxiv.org/abs/2606.08452 | https://arxiv.org/pdf/2606.08452v1 | 2606.08452 | null | 0 | 0 | false | null | null | 0.35 |
4850e5ea6ef87ce087f03ca4b0eabc6e257dbb0860acc720c50dd9907e56c299 | [
"arxiv",
"semantic_scholar"
] | Evaluating the Impact of Task Granularity on Catastrophic Forgetting in Continual Learning | Catastrophic forgetting, the abrupt loss of previously acquired knowledge upon learning new information, remains the central challenge in Continual Learning. This project investigates whether the order in which a model learns information affects how well it retains knowledge. Specifically, we ask: does learning general... | [
"Emre Alyamac",
"Himanshu Janmeda",
"Shashwat Krishna",
"Yash Vijay"
] | [
"cs.LG"
] | [
"Computer Science"
] | 2026-06-06T00:00:00 | https://arxiv.org/abs/2606.08013 | https://arxiv.org/pdf/2606.08013v1 | 2606.08013 | null | 0 | 0 | false | null | null | 0.35 |
b18fd7f43dfeabcfae6fe146458efb7e0cf2cf3e90648c5719ae460e645c15ea | [
"arxiv",
"semantic_scholar"
] | Catastrophic Forgetting as Accessibility Collapse: A Three-Level Framework for Knowledge Persistence in Continual Learning | Catastrophic forgetting is commonly interpreted as the irreversible erasure of previously acquired knowledge during sequential learning. In this work, we investigate an alternative perspective: that forgetting may arise not from complete destruction of task representations but from a loss of accessibility to preserved ... | [
"Ayushman Trivedi",
"Bhavika Melwani"
] | [
"cs.LG"
] | [
"Computer Science"
] | 2026-06-04T00:00:00 | https://arxiv.org/abs/2606.06032 | https://arxiv.org/pdf/2606.06032v1 | 2606.06032 | null | 0 | 0 | false | null | null | 0.35 |
065489ca5e818afb74c9b2672e38d9cf7e575bb2e927e5945881d4a9f3916acb | [
"arxiv",
"semantic_scholar"
] | Spurious Correlation Learning in Preference Optimization: Mechanisms, Consequences, and Mitigation via Tie Training | Preference learning methods like Direct Preference Optimization (DPO) are known to induce reliance on spurious correlations, leading to sycophancy and length bias in today's language models and potentially severe goal misgeneralization in future systems. In this work, we provide a unified theoretical analysis of this p... | [
"Christian Moya",
"Alex Semendinger",
"Guang Lin",
"Elliott Thornley"
] | [
"cs.LG",
"cs.AI"
] | [
"Computer Science"
] | 2026-05-11T00:00:00 | https://arxiv.org/abs/2605.11134 | https://arxiv.org/pdf/2605.11134v2 | 2605.11134 | null | 0 | 0 | false | null | Proceedings of the 43rd International Conference on Machine Learning, 2026, Seoul, South Korea | 0.55 |
f3e3c9c5a0c133ff30f1c3db1da3a1dc576515ff46b68e23dfa408bdaa665872 | [
"arxiv",
"semantic_scholar"
] | Overcoming Catastrophic Forgetting in Visual Continual Learning with Reinforcement Fine-Tuning | Recent studies suggest that Reinforcement Fine-Tuning (RFT) is inherently more resilient to catastrophic forgetting than Supervised Fine-Tuning (SFT). However, whether RFT (e.g., GRPO) can effectively overcome forgetting in challenging visual continual learning settings, such as class-incremental learning (CIL) and dom... | [
"Meng Lou",
"Hanzhong Guo",
"Linwei Chen",
"Yizhou Yu"
] | [
"cs.CV",
"cs.LG"
] | [
"Computer Science"
] | 2026-05-10T00:00:00 | https://arxiv.org/abs/2605.09640 | https://arxiv.org/pdf/2605.09640v1 | 2605.09640 | null | 0 | 0 | false | null | null | 0.35 |
21894e3295e39db44486c6233961e068c650f3d81d53bb682b2e40142fbcc54e | [
"arxiv",
"semantic_scholar"
] | Path-Coupled Bellman Flows for Distributional Reinforcement Learning | Distributional reinforcement learning (DRL) models the full return distribution, but existing finite-support or quantile-based methods rely on projections, while recent flow-based approaches can suffer from \emph{boundary mismatch} at the flow source or from \emph{high-variance} bootstrapping when current and successor... | [
"Boyang Xu",
"Qing Zou",
"Siqin Yang",
"Hao Yan"
] | [
"cs.LG",
"cs.AI"
] | [
"Computer Science"
] | 2026-05-07T00:00:00 | https://arxiv.org/abs/2605.08253 | https://arxiv.org/pdf/2605.08253v2 | 2605.08253 | null | 0 | 0 | false | null | Proceedings of the 43rd International Conference on Machine Learning, Seoul, South Korea. PMLR 306, 2026 | 0.55 |
4b34aaae8dfc185ede3e357ba6f020d259bba6b31b636d112040f5d3a8516493 | [
"arxiv",
"semantic_scholar"
] | Sequential Learning and Catastrophic Forgetting in Differentiable Resistor Networks | Differentiable physical networks provide a simple setting in which learning can be studied through the interaction between trainable parameters and physical equilibrium constraints. We investigate sequential learning in differentiable resistor networks governed by Kirchhoff's laws. Although individual input--output map... | [
"Maniru Ibrahim"
] | [
"cs.LG",
"cond-mat.dis-nn",
"physics.comp-ph"
] | [
"Computer Science",
"Physics"
] | 2026-05-02T00:00:00 | https://arxiv.org/abs/2605.01383 | https://arxiv.org/pdf/2605.01383v1 | 2605.01383 | null | 0 | 0 | false | null | null | 0.35 |
fe4ea4a8991e1b6e8e3d6551b34d8b952295e01c8948c3c5879fd7a201eda05d | [
"arxiv",
"semantic_scholar"
] | CI-CBM: Class-Incremental Concept Bottleneck Model for Interpretable Continual Learning | Catastrophic forgetting remains a fundamental challenge in continual learning, in which models often forget previous knowledge when fine-tuned on a new task. This issue is especially pronounced in class incremental learning (CIL), which is the most challenging setting in continual learning. Existing methods to address ... | [
"Amirhosein Javadi",
"Tuomas Oikarinen",
"Tara Javidi",
"Tsui-Wei Weng"
] | [
"cs.LG",
"cs.CV"
] | [
"Computer Science"
] | 2026-04-16T00:00:00 | https://arxiv.org/abs/2604.14519 | https://arxiv.org/pdf/2604.14519v1 | 2604.14519 | 10.48550/arXiv.2604.14519 | 0 | 0 | true | null | Transactions on Machine Learning Research, 2026 | 0.8429 |
19dfee6d24c8475c737f33849a8c8848a5e2a2c5ae71ad8a451650e14909411a | [
"arxiv",
"semantic_scholar"
] | SOLAR: A Self-Optimizing Open-Ended Autonomous Agent for Lifelong Learning and Continual Adaptation | Despite the remarkable success of large language models (LLMs), they still face bottlenecks while deploying in dynamic, real-world settings with primary challenges being concept drift and the high cost of gradient-based adaptation. Traditional fine-tuning (FT) struggles to adapt to non-stationary data streams without r... | [
"Nitin Vetcha",
"Dianbo Liu"
] | [
"cs.AI",
"cs.LG"
] | [
"Computer Science"
] | 2026-03-23T00:00:00 | https://arxiv.org/abs/2605.20189 | https://arxiv.org/pdf/2605.20189v1 | 2605.20189 | null | 0 | 0 | false | null | CEUR Workshop Proceedings, Vol. 4183, 2026 | 0.5179 |
acd3f2a5770b555b52105cb6a758171a0ead5c50e025295f6fd28fae7f0525cb | [
"arxiv",
"semantic_scholar"
] | Disentangling Dynamical Systems: Causal Representation Learning Meets Local Sparse Attention | Parametric system identification methods estimate the parameters of explicitly defined physical systems from data. Yet, they remain constrained by the need to provide an explicit function space, typically through a predefined library of candidate functions chosen via available domain knowledge. In contrast, deep learni... | [
"Markus W. Baumgartner",
"Anson Lei",
"Joe Watson",
"Ingmar Posner"
] | [
"cs.LG"
] | [
"Computer Science"
] | 2026-03-15T00:00:00 | https://arxiv.org/abs/2603.14483 | https://arxiv.org/pdf/2603.14483v2 | 2603.14483 | 10.48550/arXiv.2603.14483 | 1 | 0 | false | null | arXiv.org | 0.5088 |
e4af03ec2aafcfa998e4e76fda481799656b899df77dbce3a736bea26b63cb42 | [
"arxiv",
"semantic_scholar"
] | Chemical Reaction Networks Learn Better than Spiking Neural Networks | We mathematically prove that chemical reaction networks without hidden layers can solve tasks for which spiking neural networks require hidden layers. Our proof uses the deterministic mass-action kinetics formulation of chemical reaction networks. Specifically, we prove that a certain reaction network without hidden la... | [
"Sophie Jaffard",
"Ivo F. Sbalzarini"
] | [
"cs.LG",
"cs.AI",
"math.ST",
"stat.ML"
] | [
"Computer Science",
"Mathematics"
] | 2026-03-12T00:00:00 | https://arxiv.org/abs/2603.12060 | https://arxiv.org/pdf/2603.12060v1 | 2603.12060 | 10.48550/arXiv.2603.12060 | 0 | 0 | false | null | arXiv.org | 0.5053 |
5b57e70e78989575921a74f7098879b86f806386aaa83c2d139e79718924d0b9 | [
"arxiv",
"semantic_scholar"
] | Pretrained Vision-Language-Action Models are Surprisingly Resistant to Forgetting in Continual Learning | Continual learning is a long-standing challenge in robot policy learning, where a policy must acquire new skills over time without catastrophically forgetting previously learned ones. While prior work has extensively studied continual learning in relatively small behavior cloning (BC) policy models trained from scratch... | [
"Huihan Liu",
"Changyeon Kim",
"Bo Liu",
"Minghuan Liu",
"Yuke Zhu"
] | [
"cs.LG",
"cs.AI",
"cs.RO"
] | [
"Computer Science"
] | 2026-03-04T00:00:00 | https://arxiv.org/abs/2603.03818 | https://arxiv.org/pdf/2603.03818v2 | 2603.03818 | 10.48550/arXiv.2603.03818 | 4 | 0 | false | null | arXiv.org | 0.4961 |
b52424d3b29c08887764ee4a08292b7bd5d3e2062869471a17eb19a0adc6d7de | [
"arxiv",
"semantic_scholar"
] | Why Do Neural Networks Forget: A Study of Collapse in Continual Learning | Catastrophic forgetting is a major problem in continual learning, and lots of approaches arise to reduce it. However, most of them are evaluated through task accuracy, which ignores the internal model structure. Recent research suggests that structural collapse leads to loss of plasticity, as evidenced by changes in ef... | [
"Yunqin Zhu",
"Jun Jin"
] | [
"cs.LG",
"cs.AI"
] | [
"Computer Science"
] | 2026-03-04T00:00:00 | https://arxiv.org/abs/2603.04580 | https://arxiv.org/pdf/2603.04580v1 | 2603.04580 | 10.48550/arXiv.2603.04580 | 0 | 0 | false | null | arXiv.org | 0.4961 |
9d4dd059a647f0af15a08ed4eceebaed33f751c61372a6522ca43e5c0a77ecc2 | [
"arxiv",
"semantic_scholar"
] | Position: Modular Memory is the Key to Continual Learning Agents | Foundation models have transformed machine learning through large-scale pretraining and increased test-time compute. Despite surpassing human performance in several domains, these models remain fundamentally limited in continuous operation, experience accumulation, and personalization, capabilities that are central to ... | [
"Vaggelis Dorovatas",
"Malte Schwerin",
"Andrew D. Bagdanov",
"Lucas Caccia",
"Antonio Carta",
"Laurent Charlin",
"Barbara Hammer",
"Tyler L. Hayes",
"Timm Hess",
"Christopher Kanan",
"Dhireesha Kudithipudi",
"Xialei Liu",
"Vincenzo Lomonaco",
"Jorge Mendez-Mendez",
"Darshan Patil",
"A... | [
"cs.LG",
"cs.AI"
] | [
"Computer Science"
] | 2026-03-02T00:00:00 | https://arxiv.org/abs/2603.01761 | https://arxiv.org/pdf/2603.01761v2 | 2603.01761 | 10.48550/arXiv.2603.01761 | 1 | 0 | false | null | arXiv.org | 0.4939 |
1e1560bb31fac79cd470c3115d6885bf4ce92f1e1b2d538f66384209632be9bd | [
"arxiv",
"semantic_scholar"
] | Learning in the Null Space: Small Singular Values for Continual Learning | Alleviating catastrophic forgetting while enabling further learning is a primary challenge in continual learning (CL). Orthogonal-based training methods have gained attention for their efficiency and strong theoretical properties, and many existing approaches enforce orthogonality through gradient projection. In this p... | [
"Cuong Anh Pham",
"Praneeth Vepakomma",
"Samuel HorvΓ‘th"
] | [
"cs.LG",
"cs.CV"
] | [
"Computer Science"
] | 2026-02-25T00:00:00 | https://arxiv.org/abs/2602.21919 | https://arxiv.org/pdf/2602.21919v1 | 2602.21919 | 10.48550/arXiv.2602.21919 | 0 | 0 | true | https://github.com/pacman-ctm/NESS | arXiv.org | 0.7544 |
17d5f96ef84969f5a386fee224cdef9af4ecb2612d7d9e0a9c31a0fc1dd987b1 | [
"arxiv",
"semantic_scholar"
] | Value Bonuses using Ensemble Errors for Exploration in Reinforcement Learning | Optimistic value estimates provide one mechanism for directed exploration in reinforcement learning (RL). The agent acts greedily with respect to an estimate of the value plus what can be seen as a value bonus. The value bonus can be learned by estimating a value function on reward bonuses, propagating local uncertaint... | [
"Abdul Wahab",
"Raksha Kumaraswamy",
"Martha White"
] | [
"cs.LG",
"cs.AI"
] | [
"Computer Science"
] | 2026-02-12T00:00:00 | https://arxiv.org/abs/2602.12375 | https://arxiv.org/pdf/2602.12375v1 | 2602.12375 | 10.48550/arXiv.2602.12375 | 0 | 0 | false | null | arXiv.org | 0.4732 |
d790c88703165f0f1bb398150b9539961ceca2247724ee68a9c4f55d5239f3bd | [
"arxiv",
"semantic_scholar"
] | MerLin: A Discovery Engine for Photonic and Hybrid Quantum Machine Learning | Identifying where quantum models may offer practical benefits in near term quantum machine learning (QML) requires moving beyond isolated algorithmic proposals toward systematic and empirical exploration across models, datasets, and hardware constraints. We introduce MerLin, an open-source framework designed as a disco... | [
"Cassandre Notton",
"Benjamin Stott",
"Philippe Schoeb",
"Anthony Walsh",
"GrΓ©goire Leboucher",
"Vincent Espitalier",
"Vassilis Apostolou",
"Louis-FΓ©lix Vigneux",
"Alexia Salavrakos",
"Jean Senellart"
] | [
"cs.LG",
"cs.PL",
"quant-ph"
] | [
"Computer Science",
"Physics"
] | 2026-02-11T00:00:00 | https://arxiv.org/abs/2602.11092 | https://arxiv.org/pdf/2602.11092v2 | 2602.11092 | 10.48550/arXiv.2602.11092 | 1 | 0 | true | null | arXiv.org | 0.7296 |
0d014a287f0d92d1d96ea9be47e66efbad5a22a27255bed97f6e39c24a437ac4 | [
"arxiv",
"semantic_scholar"
] | A Thermodynamic Theory of Learning Part II: Critical Period Closure and Continual Learning Failure | Learning performed over finite time is inherently irreversible. In Part~I of this series, we modeled learning as a transport process in the space of parameter distributions and derived the Epistemic Speed Limit (ESL), which lower-bounds entropy production under finite-time dynamics. In this work (Part~II), we show that... | [
"Daisuke Okanohara"
] | [
"cs.LG"
] | [
"Computer Science"
] | 2026-02-08T00:00:00 | https://arxiv.org/abs/2602.07950 | https://arxiv.org/pdf/2602.07950v2 | 2602.07950 | 10.48550/arXiv.2602.07950 | 0 | 0 | false | null | arXiv.org | 0.4686 |
8e0f0b27f0ed9c58c12203cab5c956ac0cc9f12460329a9b1ff2499be9353e3a | [
"arxiv",
"semantic_scholar"
] | Keep Rehearsing and Refining: Lifelong Learning Vehicle Routing under Continually Drifting Tasks | Existing neural solvers for vehicle routing problems (VRPs) are typically trained either in a one-off manner on a fixed set of pre-defined tasks or in a lifelong manner with tasks arriving sequentially, assuming sufficient training on each task. Both settings overlook a common real-world property: problem patterns may ... | [
"Jiyuan Pei",
"Yi Mei",
"Jialin Liu",
"Mengjie Zhang",
"Xin Yao"
] | [
"cs.LG",
"cs.AI"
] | [
"Computer Science"
] | 2026-01-30T00:00:00 | https://arxiv.org/abs/2601.22509 | https://arxiv.org/pdf/2601.22509v2 | 2601.22509 | 10.48550/arXiv.2601.22509 | 0 | 0 | false | null | arXiv.org | 0.4583 |
7e6a12b9aa29842f32aef94bba1255f3e4f70ecfa69a5647e64d254b6f339d22 | [
"arxiv",
"semantic_scholar"
] | Federated Learning Under Temporal Drift -- Mitigating Catastrophic Forgetting via Experience Replay | Federated Learning struggles under temporal concept drift where client data distributions shift over time. We demonstrate that standard FedAvg suffers catastrophic forgetting under seasonal drift on Fashion-MNIST, with accuracy dropping from 74% to 28%. We propose client-side experience replay, where each client mainta... | [
"Sahasra Kokkula",
"Daniel David",
"Aaditya Baruah"
] | [
"cs.LG",
"cs.DC"
] | [
"Computer Science"
] | 2026-01-19T00:00:00 | https://arxiv.org/abs/2601.13456 | https://arxiv.org/pdf/2601.13456v1 | 2601.13456 | 10.48550/arXiv.2601.13456 | 0 | 0 | false | null | arXiv.org | 0.4457 |
5914cf9b44fd29434dcce3c96c191fc84d4853b285a8bb4fafaced78ba3750d8 | [
"arxiv",
"semantic_scholar"
] | Exploring Student Expectations and Confidence in Learning Analytics | Learning Analytics (LA) is nowadays ubiquitous in many educational systems, providing the ability to collect and analyze student data in order to understand and optimize learning and the environments in which it occurs. On the other hand, the collection of data requires to comply with the growing demand regarding priva... | [
"Hayk Asatryan",
"Basile Tousside",
"Janis Mohr",
"Malte Neugebauer",
"Hildo Bijl",
"Paul Spiegelberg",
"Claudia Frohn-Schauf",
"JΓΆrg Frochte"
] | [
"cs.LG",
"cs.CY",
"cs.HC"
] | [
"Computer Science"
] | 2026-01-08T00:00:00 | https://arxiv.org/abs/2601.05082 | https://arxiv.org/pdf/2601.05082v1 | 2601.05082 | 10.1145/3636555.3636923 | 2 | 0 | false | null | International Conference on Learning Analytics and Knowledge | 0.4331 |
618492aa2a53a0f44392593bafa31587bdcec2b909b87ab09332a63b00b16651 | [
"arxiv",
"semantic_scholar"
] | Shallow Neural Networks Learn Low-Degree Spherical Polynomials with Feature Learning by Learnable Channel Attention | We study the problem of learning a low-degree spherical polynomial of degree $\ell_0 = Ξ(1) \ge 1$ defined on the unit sphere in $\RR^d$ by training an over-parameterized two-layer neural network (NN) with channel attention in this paper. Our main result is the significantly improved sample complexity for learning such... | [
"Yingzhen Yang"
] | [
"stat.ML",
"cs.LG",
"math.OC"
] | [
"Mathematics",
"Computer Science"
] | 2025-12-23T00:00:00 | https://arxiv.org/abs/2512.20562 | https://arxiv.org/pdf/2512.20562v2 | 2512.20562 | null | 0 | 0 | false | null | null | 0.264 |
74eaccfd7edd9bf62a60b0f99b0db8b1cd74e122db4d24887b93c9a6a9ede1d7 | [
"arxiv",
"semantic_scholar"
] | Sequencing to Mitigate Catastrophic Forgetting in Continual Learning | To cope with real-world dynamics, an intelligent system needs to incrementally acquire, update, and exploit knowledge throughout its lifetime. This ability, known as Continual learning, provides a foundation for AI systems to develop themselves adaptively. Catastrophic forgetting is a major challenge to the progress of... | [
"Hesham G. Moussa",
"Aroosa Hameed",
"Arashmid Akhavain"
] | [
"cs.LG",
"cs.AI"
] | [
"Computer Science"
] | 2025-12-18T00:00:00 | https://arxiv.org/abs/2512.16871 | https://arxiv.org/pdf/2512.16871v1 | 2512.16871 | 10.48550/arXiv.2512.16871 | 0 | 0 | false | null | arXiv.org | 0.4091 |
78d4cae4f641146f4bbd6c08318b815ca2bd8780258e1a2a811ff630a80cd334 | [
"arxiv",
"semantic_scholar"
] | Continual Learning at the Edge: An Agnostic IIoT Architecture | The exponential growth of Internet-connected devices has presented challenges to traditional centralized computing systems due to latency and bandwidth limitations. Edge computing has evolved to address these difficulties by bringing computations closer to the data source. Additionally, traditional machine learning alg... | [
"Pablo GarcΓa-Santaclara",
"Bruno FernΓ‘ndez-Castro",
"Rebeca P. DΓaz-Redondo",
"Carlos Calvo-Moa",
"Henar MariΓ±o-BodelΓ³n"
] | [
"stat.ML",
"cs.LG"
] | [
"Computer Science",
"Mathematics"
] | 2025-12-16T00:00:00 | https://arxiv.org/abs/2512.14311 | https://arxiv.org/pdf/2512.14311v1 | 2512.14311 | 10.1007/978-981-96-6938-7_33 | 1 | 0 | false | null | arXiv.org | 0.4068 |
14b4c3e551bff6b7a8eca4df632d968f6504f9170ef77f3b0f3c36ff7e3ad8d9 | [
"arxiv",
"semantic_scholar"
] | Multiclass Graph-Based Large Margin Classifiers: Unified Approach for Support Vectors and Neural Networks | While large margin classifiers are originally an outcome of an optimization framework, support vectors (SVs) can be obtained from geometric approaches. This article presents advances in the use of Gabriel graphs (GGs) in binary and multiclass classification problems. For Chipclass, a hyperparameter-less and optimizatio... | [
"VΓtor M. Hanriot",
"Luiz C. B. Torres",
"AntΓ΄nio P. Braga"
] | [
"cs.LG",
"stat.ML"
] | [
"Computer Science",
"Medicine",
"Mathematics"
] | 2025-12-15T00:00:00 | https://arxiv.org/abs/2512.13410 | https://arxiv.org/pdf/2512.13410v1 | 2512.13410 | 10.1109/TNNLS.2024.3420227 | 3 | 1 | false | null | IEEE Transactions on Neural Networks and Learning Systems | 0.4056 |
455a5c2cb4fe3c769f06beaf1267f20415aff21f777759daf1a0dbfb87c43a0f | [
"arxiv",
"semantic_scholar"
] | Bridging Streaming Continual Learning via In-Context Large Tabular Models | In streaming scenarios, models must learn continuously, adapting to concept drifts without erasing previously acquired knowledge. However, existing research communities address these challenges in isolation. Continual Learning (CL) focuses on long-term retention and mitigating catastrophic forgetting, often without str... | [
"Afonso Lourenço",
"JoΓ£o Gama",
"Eric P. Xing",
"Goreti Marreiros"
] | [
"cs.LG"
] | [
"Computer Science"
] | 2025-12-12T00:00:00 | https://arxiv.org/abs/2512.11668 | https://arxiv.org/pdf/2512.11668v1 | 2512.11668 | 10.48550/arXiv.2512.11668 | 4 | 0 | false | null | null | 0.2559 |
4bb3a6793a83c3c7835a15ba82993513caa574d4efea3991530e94b926a5a78d | [
"arxiv",
"semantic_scholar"
] | Angular Regularization for Positive-Unlabeled Learning on the Hypersphere | Positive-Unlabeled (PU) learning addresses classification problems where only a subset of positive examples is labeled and the remaining data is unlabeled, making explicit negative supervision unavailable. Existing PU methods often rely on negative-risk estimation or pseudo-labeling, which either require strong distrib... | [
"Vasileios Sevetlidis",
"George Pavlidis",
"Antonios Gasteratos"
] | [
"cs.LG",
"cs.AI"
] | [
"Computer Science"
] | 2025-12-07T00:00:00 | https://arxiv.org/abs/2512.06785 | https://arxiv.org/pdf/2512.06785v1 | 2512.06785 | 10.48550/arXiv.2512.06785 | 1 | 0 | false | null | Transactions on Machine Learning Research, 2025 | 0.3965 |
09e986ece22d720241038e4e1760c1feab3bd4a18b9d0b2e58f3d5d6c84ea75d | [
"arxiv",
"semantic_scholar"
] | Mitigating Catastrophic Forgetting in Mathematical Reasoning Finetuning through Mixed Training | When finetuning large language models for specialized tasks such as mathematical reasoning, models exhibit catastrophic forgetting, losing previously learned capabilities. We investigate this by finetuning Flan-T5-Base (250M parameters) on the DeepMind Mathematics dataset and measuring forgetting on MultiNLI. Math-only... | [
"John Graham Reynolds"
] | [
"cs.LG",
"cs.CL"
] | [
"Computer Science"
] | 2025-12-05T00:00:00 | https://arxiv.org/abs/2512.13706 | https://arxiv.org/pdf/2512.13706v1 | 2512.13706 | 10.48550/arXiv.2512.13706 | 0 | 0 | true | https://github.com/johngrahamreynolds/mathematical_catastrophe_mitigation | arXiv.org | 0.6092 |
73f4e5ea821d529d5ee0b07719a06076fe00b43a43629f1eb7599a437d2766f6 | [
"arxiv",
"semantic_scholar"
] | Sample Complexity of Distributionally Robust Off-Dynamics Reinforcement Learning with Online Interaction | Off-dynamics reinforcement learning (RL), where training and deployment transition dynamics are different, can be formulated as learning in a robust Markov decision process (RMDP) where uncertainties in transition dynamics are imposed. Existing literature mostly assumes access to generative models allowing arbitrary st... | [
"Yiting He",
"Zhishuai Liu",
"Weixin Wang",
"Pan Xu"
] | [
"cs.LG",
"cs.AI",
"cs.RO",
"stat.ML"
] | [
"Computer Science",
"Mathematics"
] | 2025-11-07T00:00:00 | https://arxiv.org/abs/2511.05396 | https://arxiv.org/pdf/2511.05396v1 | 2511.05396 | 10.48550/arXiv.2511.05396 | 13 | 1 | false | null | International Conference on Machine Learning | 0.3621 |
4f6770d70b203cd13be6e63f9ff466ab2a3e1de424619d79997453fe6709ce19 | [
"arxiv",
"semantic_scholar"
] | Forgetting is Everywhere | A fundamental challenge in developing general learning algorithms is their tendency to forget past knowledge when adapting to new data. Addressing this problem requires a principled understanding of forgetting; yet, despite decades of study, no unified definition has emerged that provides insights into the underlying d... | [
"Ben Sanati",
"Thomas L. Lee",
"Trevor McInroe",
"Aidan Scannell",
"Nikolay Malkin",
"David Abel",
"Amos Storkey"
] | [
"cs.LG",
"stat.ML"
] | [
"Computer Science",
"Mathematics"
] | 2025-11-06T00:00:00 | https://arxiv.org/abs/2511.04666 | https://arxiv.org/pdf/2511.04666v3 | 2511.04666 | 10.48550/arXiv.2511.04666 | 0 | 0 | false | null | arXiv.org | 0.3609 |
78465d6e36b164bbf07bf6119676da7d0c8dee742fe618e5cb8208a674b31f5d | [
"arxiv",
"semantic_scholar"
] | Benchmarking Catastrophic Forgetting Mitigation Methods in Federated Time Series Forecasting | Catastrophic forgetting (CF) poses a persistent challenge in continual learning (CL), especially within federated learning (FL) environments characterized by non-i.i.d. time series data. While existing research has largely focused on classification tasks in vision domains, the regression-based forecasting setting preva... | [
"Khaled Hallak",
"Oudom Kem"
] | [
"cs.LG",
"cs.DC",
"stat.ML"
] | [
"Computer Science",
"Mathematics"
] | 2025-10-24T00:00:00 | https://arxiv.org/abs/2510.21491 | https://arxiv.org/pdf/2510.21491v1 | 2510.21491 | 10.1109/FLTA67013.2025.11336525 | 0 | 0 | true | null | null | 0.409 |
24dce3c2adbb9c65aaa655e01bffd3d5f34dc85f2ff8b833915c0620a1da29c2 | [
"arxiv",
"semantic_scholar"
] | Multi-modal Co-learning for Earth Observation: Enhancing single-modality models via modality collaboration | Multi-modal co-learning is emerging as an effective paradigm in machine learning, enabling models to collaboratively learn from different modalities to enhance single-modality predictions. Earth Observation (EO) represents a quintessential domain for multi-modal data analysis, wherein diverse remote sensors collect dat... | [
"Francisco Mena",
"Dino Ienco",
"Cassio F. Dantas",
"Roberto Interdonato",
"Andreas Dengel"
] | [
"cs.CV",
"cs.AI",
"cs.LG"
] | [
"Computer Science"
] | 2025-10-22T00:00:00 | https://arxiv.org/abs/2510.19579 | https://arxiv.org/pdf/2510.19579v1 | 2510.19579 | 10.1007/s10994-025-06903-0 | 3 | 0 | false | null | Machine-mediated learning | 0.3438 |
c9c31f691994cefa3ff3ab942e99940a4c7d6887f166245009e86162c9616008 | [
"arxiv",
"semantic_scholar"
] | On the Implicit Adversariality of Catastrophic Forgetting in Deep Continual Learning | Continual learning seeks the human-like ability to accumulate new skills in machine intelligence. Its central challenge is catastrophic forgetting, whose underlying cause has not been fully understood for deep networks. In this paper, we demystify catastrophic forgetting by revealing that the new-task training is impli... | [
"Ze Peng",
"Jian Zhang",
"Jintao Guo",
"Lei Qi",
"Yang Gao",
"Yinghuan Shi"
] | [
"cs.LG",
"cs.AI"
] | [
"Computer Science"
] | 2025-10-10T00:00:00 | https://arxiv.org/abs/2510.09181 | https://arxiv.org/pdf/2510.09181v1 | 2510.09181 | 10.48550/arXiv.2510.09181 | 0 | 0 | false | null | arXiv.org | 0.33 |
663a026587d5a7fc9cffbbe310958f18dd9c73027725b813c7c3de8b4af82c1f | [
"arxiv",
"semantic_scholar"
] | On the Theory of Continual Learning with Gradient Descent for Neural Networks | Continual learning, the ability of a model to adapt to an ongoing sequence of tasks without forgetting earlier ones, is a central goal of artificial intelligence. To better understand its underlying mechanisms, we study the limitations of continual learning in a tractable yet representative setting. Specifically, we an... | [
"Hossein Taheri",
"Avishek Ghosh",
"Arya Mazumdar"
] | [
"stat.ML",
"cs.IT",
"cs.LG"
] | [
"Mathematics",
"Computer Science"
] | 2025-10-07T00:00:00 | https://arxiv.org/abs/2510.05573 | https://arxiv.org/pdf/2510.05573v2 | 2510.05573 | 10.48550/arXiv.2510.05573 | 0 | 0 | false | null | arXiv.org | 0.3266 |
97b8b6b1386b59c7ee8404c91f2ce44e69791866e8d2225c6bb562d2599b411b | [
"arxiv",
"semantic_scholar"
] | Learning Time-Series Representations by Hierarchical Uniformity-Tolerance Latent Balancing | We propose TimeHUT, a novel method for learning time-series representations by hierarchical uniformity-tolerance balancing of contrastive representations. Our method uses two distinct losses to learn strong representations with the aim of striking an effective balance between uniformity and tolerance in the embedding s... | [
"Amin Jalali",
"Milad Soltany",
"Michael Greenspan",
"Ali Etemad"
] | [
"cs.LG",
"cs.AI"
] | [
"Computer Science"
] | 2025-10-02T00:00:00 | https://arxiv.org/abs/2510.01658 | https://arxiv.org/pdf/2510.01658v1 | 2510.01658 | 10.48550/arXiv.2510.01658 | 1 | 0 | false | null | Transactions on Machine Learning Research (10/2025) | 0.3208 |
3ead865678e382e9f9745cbc75e915bc4cf93e0e51bac3f20204cdb5e4dfaad8 | [
"arxiv",
"semantic_scholar"
] | Generative Evolutionary Meta-Solver (GEMS): Scalable Surrogate-Free Multi-Agent Reinforcement Learning | Scalable multi-agent reinforcement learning (MARL) remains a central challenge for AI. Existing population-based methods, like Policy-Space Response Oracles, PSRO, require storing explicit policy populations and constructing full payoff matrices, incurring quadratic computation and linear memory costs. We present Gener... | [
"Alakh Sharma",
"Gaurish Trivedi",
"Kartikey Singh Bhandari",
"Yash Sinha",
"Dhruv Kumar",
"Pratik Narang",
"Jagat Sesh Challa"
] | [
"cs.LG",
"cs.AI"
] | [
"Computer Science"
] | 2025-09-27T00:00:00 | https://arxiv.org/abs/2509.23462 | https://arxiv.org/pdf/2509.23462v2 | 2509.23462 | null | 0 | 0 | false | null | Transactions on Machine Learning Research (2026) | 0.3151 |
8d027c3ac314a66e4df680ccfab767547c7ad7673374c1acec3313d58b38dfca | [
"arxiv",
"semantic_scholar"
] | Lifelong Learning with Behavior Consolidation for Vehicle Routing | Recent neural solvers have demonstrated promising performance in learning to solve routing problems. However, existing studies are primarily based on one-off training on one or a set of predefined problem distributions and scales, i.e., tasks. When a new task arises, they typically rely on either zero-shot generalizati... | [
"Jiyuan Pei",
"Yi Mei",
"Jialin Liu",
"Mengjie Zhang",
"Xin Yao"
] | [
"cs.AI",
"cs.LG"
] | [
"Computer Science"
] | 2025-09-26T00:00:00 | https://arxiv.org/abs/2509.21765 | https://arxiv.org/pdf/2509.21765v4 | 2509.21765 | 10.48550/arXiv.2509.21765 | 1 | 0 | false | null | arXiv.org | 0.314 |
c0b01bdc837d0ea3c43cf0a3d646a3023451b12ac9df0fdbfae6e0b0a1109e02 | [
"arxiv",
"semantic_scholar"
] | Intra-Cluster Mixup: An Effective Data Augmentation Technique for Complementary-Label Learning | In this paper, we investigate the challenges of complementary-label learning (CLL), a specialized form of weakly-supervised learning (WSL) where models are trained with labels indicating classes to which instances do not belong, rather than standard ordinary labels. This alternative supervision is appealing because col... | [
"Tan-Ha Mai",
"Hsuan-Tien Lin"
] | [
"cs.LG",
"cs.AI",
"cs.CV"
] | [
"Computer Science"
] | 2025-09-22T00:00:00 | https://arxiv.org/abs/2509.17971 | https://arxiv.org/pdf/2509.17971v2 | 2509.17971 | 10.48550/arXiv.2509.17971 | 3 | 0 | false | null | Transactions on Machine Learning Research, 2026 | 0.3094 |
964fb412be437914c7fa790ee74f680456650d4f4cb6df184347e3b1b007478c | [
"arxiv",
"semantic_scholar"
] | Fourier Learning Machines: Nonharmonic Fourier-Based Neural Networks for Scientific Machine Learning | We introduce the Fourier Learning Machine (FLM), a neural network (NN) architecture designed to represent a multidimensional nonharmonic Fourier series. The FLM uses a simple feedforward structure with cosine activation functions to learn the frequencies, amplitudes, and phase shifts of the series as trainable paramete... | [
"Mominul Rubel",
"Adam Meyers",
"Gabriel Nicolosi"
] | [
"cs.LG",
"math.OC"
] | [
"Computer Science",
"Mathematics"
] | 2025-09-10T00:00:00 | https://arxiv.org/abs/2509.08759 | https://arxiv.org/pdf/2509.08759v3 | 2509.08759 | 10.48550/arXiv.2509.08759 | 1 | 0 | false | null | Transactions on Machine Learning Research, December 2025 | 0.2956 |
925621ad2f75055c0e13214a0c6a4927b140bf9f4df4427650851e72e6d8d93f | [
"arxiv",
"semantic_scholar"
] | Gaming and Cooperation in Federated Learning: What Can Happen and How to Monitor It | The success of federated learning (FL) ultimately depends on how strategic participants behave under partial observability, yet most formulations still treat FL as a static optimization problem. We instead view FL deployments as governed strategic systems and develop an analytical framework that separates welfare-impro... | [
"Dongseok Kim",
"Hyoungsun Choi",
"Mohamed Jismy Aashik Rasool",
"Gisung Oh"
] | [
"cs.LG",
"cs.GT",
"stat.ML"
] | [
"Computer Science",
"Mathematics"
] | 2025-09-02T00:00:00 | https://arxiv.org/abs/2509.02391 | https://arxiv.org/pdf/2509.02391v3 | 2509.02391 | 10.48550/arXiv.2509.02391 | 0 | 0 | false | null | Transactions on Machine Learning Research, 2026 | 0.2865 |
860d720d28936d99dcf2e9dcf6fc70654482885d2bbdca6703da4aafb23f686b | [
"arxiv",
"semantic_scholar"
] | Mitigating Catastrophic Forgetting in Continual Learning through Model Growth | Catastrophic forgetting is a significant challenge in continual learning, in which a model loses prior knowledge when it is fine-tuned on new tasks. This problem is particularly critical for large language models (LLMs) undergoing continual learning, as retaining performance across diverse domains is important for thei... | [
"Ege SΓΌalp",
"Mina Rezaei"
] | [
"cs.CL"
] | [
"Computer Science"
] | 2025-09-01T00:00:00 | https://arxiv.org/abs/2509.01213 | https://arxiv.org/pdf/2509.01213v1 | 2509.01213 | 10.48550/arXiv.2509.01213 | 3 | 0 | false | null | arXiv.org | 0.2853 |
2e50390f1d586dddb134ce20edec8b8e4efa39f1e332949a87e8a6e7b2a72b42 | [
"arxiv",
"semantic_scholar"
] | Online Learning with Multiple Fairness Regularizers via Graph-Structured Feedback | There is an increasing need to enforce multiple, often competing, measures of fairness within automated decision systems. The appropriate weighting of these fairness objectives is typically unknown a priori, may change over time and, in our setting, must be learned adaptively through sequential interactions. In this wo... | [
"Quan Zhou",
"Jakub Marecek",
"Robert Shorten"
] | [
"cs.LG",
"cs.AI"
] | [
"Computer Science"
] | 2025-08-19T00:00:00 | https://arxiv.org/abs/2508.14311 | https://arxiv.org/pdf/2508.14311v2 | 2508.14311 | null | 0 | 0 | false | null | Transactions on Machine Learning Research (TMLR), 2026 | 0.2704 |
80cb07b9ce32641a3efa439619a295629fae2fb8c39dfed4162ade58bce5f004 | [
"arxiv",
"semantic_scholar"
] | H2C: Hippocampal Circuit-inspired Continual Learning for Lifelong Trajectory Prediction in Autonomous Driving | Deep learning (DL) has shown state-of-the-art performance in trajectory prediction, which is critical to safe navigation in autonomous driving (AD). However, most DL-based methods suffer from catastrophic forgetting, where adapting to a new distribution may cause significant performance degradation in previously learne... | [
"Yunlong Lin",
"Zirui Li",
"Guodong Du",
"Xiaocong Zhao",
"Cheng Gong",
"Xinwei Wang",
"Chao Lu",
"Jianwei Gong"
] | [
"cs.AI"
] | [
"Computer Science"
] | 2025-08-02T00:00:00 | https://arxiv.org/abs/2508.01158 | https://arxiv.org/pdf/2508.01158v2 | 2508.01158 | 10.48550/arXiv.2508.01158 | 2 | 0 | true | https://github.com/BIT-Jack/H2C-lifelong | arXiv.org | 0.3878 |
68f146317098e49864d32d0f011275318d5bdfd964cf3ffaf3c68e25bc2d992f | [
"arxiv"
] | Continual Generalized Category Discovery: Learning and Forgetting from a Bayesian Perspective | Continual Generalized Category Discovery (C-GCD) faces a critical challenge: incrementally learning new classes from unlabeled data streams while preserving knowledge of old classes. Existing methods struggle with catastrophic forgetting, especially when unlabeled data mixes known and novel categories. We address this ... | [
"Hao Dai",
"Jagmohan Chauhan"
] | [
"cs.LG"
] | [] | 2025-07-23T00:00:00 | https://arxiv.org/abs/2507.17382 | https://arxiv.org/pdf/2507.17382v1 | 2507.17382 | null | 0 | 0 | true | https://github.com/daihao42/VB-CGCD | null | 0.283 |
102542a239d1825abc93b95b2bdd95e6d6a7c647e9e911e610a115e133061965 | [
"arxiv",
"semantic_scholar"
] | Overcoming catastrophic forgetting in neural networks | Catastrophic forgetting is the primary challenge that hinders continual learning, which refers to a neural network ability to sequentially learn multiple tasks while retaining previously acquired knowledge. Elastic Weight Consolidation, a regularization-based approach inspired by synaptic consolidation in biological ne... | [
"Brandon Shuen Yi Loke",
"Filippo Quadri",
"Gabriel Vivanco",
"Maximilian Casagrande",
"SaΓΊl Fenollosa"
] | [
"cs.LG",
"cs.IR"
] | [
"Computer Science"
] | 2025-07-14T00:00:00 | https://arxiv.org/abs/2507.10485 | https://arxiv.org/pdf/2507.10485v1 | 2507.10485 | 10.48550/arXiv.2507.10485 | 160 | 15 | false | null | arXiv.org | 0.6021 |
e47a0f50f2fc3ed29368a843d20972839f28346eb9589f15720da6ed00d18e44 | [
"arxiv",
"semantic_scholar"
] | How Weight Resampling and Optimizers Shape the Dynamics of Continual Learning and Forgetting in Neural Networks | Recent work in continual learning has highlighted the beneficial effect of resampling weights in the last layer of a neural network (``zapping"). Although empirical results demonstrate the effectiveness of this approach, the underlying mechanisms that drive these improvements remain unclear. In this work, we investigat... | [
"Lapo Frati",
"Neil Traft",
"Jeff Clune",
"Nick Cheney"
] | [
"cs.LG",
"cs.CV"
] | [
"Computer Science"
] | 2025-07-02T00:00:00 | https://arxiv.org/abs/2507.01559 | https://arxiv.org/pdf/2507.01559v1 | 2507.01559 | 10.48550/arXiv.2507.01559 | 0 | 0 | false | null | arXiv.org | 0.2154 |
03fdadd84be14b67b4e959afeac998c72626d098a0e1ab57b5bca6f061c6cdee | [
"arxiv",
"semantic_scholar"
] | The Importance of Being Lazy: Scaling Limits of Continual Learning | Despite recent efforts, neural networks still struggle to learn in non-stationary environments, and our understanding of catastrophic forgetting (CF) is far from complete. In this work, we perform a systematic study on the impact of model scale and the degree of feature learning in continual learning. We reconcile exis... | [
"Jacopo Graldi",
"Alessandro Breccia",
"Giulia Lanzillotta",
"Thomas Hofmann",
"Lorenzo Noci"
] | [
"cs.LG",
"cs.AI",
"stat.ML"
] | [
"Computer Science",
"Mathematics"
] | 2025-06-20T00:00:00 | https://arxiv.org/abs/2506.16884 | https://arxiv.org/pdf/2506.16884v2 | 2506.16884 | 10.48550/arXiv.2506.16884 | 4 | 0 | false | null | International Conference on Machine Learning | 0.2017 |
d627c7ca37ba34f90c9a0f348aacab5831e1e03bf382e25919719546b59be663 | [
"arxiv",
"semantic_scholar"
] | FedOne: Query-Efficient Federated Learning for Black-box Discrete Prompt Learning | Black-Box Discrete Prompt Learning is a prompt-tuning method that optimizes discrete prompts without accessing model parameters or gradients, making the prompt tuning on a cloud-based Large Language Model (LLM) feasible. Adapting federated learning to BDPL could further enhance prompt tuning performance by leveraging d... | [
"Ganyu Wang",
"Jinjie Fang",
"Maxwell J. Yin",
"Bin Gu",
"Xi Chen",
"Boyu Wang",
"Yi Chang",
"Charles Ling"
] | [
"cs.LG"
] | [
"Computer Science"
] | 2025-06-17T00:00:00 | https://arxiv.org/abs/2506.14929 | https://arxiv.org/pdf/2506.14929v2 | 2506.14929 | 10.48550/arXiv.2506.14929 | 2 | 0 | false | null | International Conference on Machine Learning | 0.1982 |
a4cb0b70b24b402d7546f3c9179fccf00cf87b055777e92d42484473cb17dad8 | [
"arxiv",
"semantic_scholar"
] | On the Similarities of Embeddings in Contrastive Learning | Contrastive learning operates on a simple yet effective principle: Embeddings of positive pairs are pulled together, while those of negative pairs are pushed apart. In this paper, we propose a unified framework for understanding contrastive learning through the lens of cosine similarity, and present two key theoretical... | [
"Chungpa Lee",
"Sehee Lim",
"Kibok Lee",
"Jy-yong Sohn"
] | [
"cs.LG",
"stat.ML"
] | [
"Computer Science",
"Mathematics"
] | 2025-06-11T00:00:00 | https://arxiv.org/abs/2506.09781 | https://arxiv.org/pdf/2506.09781v2 | 2506.09781 | 10.48550/arXiv.2506.09781 | 2 | 0 | false | null | International Conference on Machine Learning | 0.1914 |
ee2b08b75c51e46f4b7635e1253d85ca1a18c521036a319853dd7eaeed36be67 | [
"arxiv",
"semantic_scholar"
] | Federated Learning on Stochastic Neural Networks | Federated learning is a machine learning paradigm that leverages edge computing on client devices to optimize models while maintaining user privacy by ensuring that local data remains on the device. However, since all data is collected by clients, federated learning is susceptible to latent noise in local datasets. Fac... | [
"Jingqiao Tang",
"Ryan Bausback",
"Feng Bao",
"Richard Archibald"
] | [
"cs.LG",
"cs.DC"
] | [
"Computer Science"
] | 2025-06-09T00:00:00 | https://arxiv.org/abs/2506.08169 | https://arxiv.org/pdf/2506.08169v1 | 2506.08169 | 10.48550/arXiv.2506.08169 | 1 | 0 | false | null | arXiv.org | 0.1891 |
a7214f4049d0a30f227516276a52f1b66537e05bea341c926caf35214051b115 | [
"arxiv",
"semantic_scholar"
] | Dynamic Mixture of Progressive Parameter-Efficient Expert Library for Lifelong Robot Learning | A generalist agent must continuously learn and adapt throughout its lifetime, achieving efficient forward transfer while minimizing catastrophic forgetting. Previous work within the dominant pretrain-then-finetune paradigm has explored parameter-efficient fine-tuning for single-task adaptation, effectively steering a f... | [
"Yuheng Lei",
"Sitong Mao",
"Shunbo Zhou",
"Hongyuan Zhang",
"Xuelong Li",
"Ping Luo"
] | [
"cs.LG",
"cs.RO"
] | [
"Computer Science"
] | 2025-06-06T00:00:00 | https://arxiv.org/abs/2506.05985 | https://arxiv.org/pdf/2506.05985v3 | 2506.05985 | 10.48550/arXiv.2506.05985 | 4 | 0 | true | https://github.com/HarryLui98/DMPEL | arXiv.org | 0.2869 |
09a53dbe215f894824ca223cd204aae1ddcbe2f28395cfb2c40bf3b20c00a31d | [
"arxiv",
"semantic_scholar"
] | Can LLMs Alleviate Catastrophic Forgetting in Graph Continual Learning? A Systematic Study | Nowadays, real-world data, including graph-structure data, often arrives in a streaming manner, which means that learning systems need to continuously acquire new knowledge without forgetting previously learned information. Although substantial existing works attempt to address catastrophic forgetting in graph machine ... | [
"Ziyang Cheng",
"Zhixun Li",
"Yuhan Li",
"Yixin Song",
"Kangyi Zhao",
"Dawei Cheng",
"Jia Li",
"Hong Cheng",
"Jeffrey Xu Yu"
] | [
"cs.LG",
"cs.AI"
] | [
"Computer Science"
] | 2025-05-24T00:00:00 | https://arxiv.org/abs/2505.18697 | https://arxiv.org/pdf/2505.18697v2 | 2505.18697 | 10.48550/arXiv.2505.18697 | 3 | 0 | true | https://github.com/ZhixunLEE/LLM4GCL | arXiv.org | 0.2639 |
fd254efb1bf0f3d7f39e441cce8586c91a4c54529e5cd8531c27a281506ab490 | [
"arxiv",
"semantic_scholar"
] | Efficient Training of Neural SDEs Using Stochastic Optimal Control | We present a hierarchical, control theory inspired method for variational inference (VI) for neural stochastic differential equations (SDEs). While VI for neural SDEs is a promising avenue for uncertainty-aware reasoning in time-series, it is computationally challenging due to the iterative nature of maximizing the ELB... | [
"Rembert Daems",
"Manfred Opper",
"Guillaume Crevecoeur",
"Tolga Birdal"
] | [
"cs.LG",
"cs.AI",
"math.PR"
] | [
"Computer Science",
"Mathematics"
] | 2025-05-22T00:00:00 | https://arxiv.org/abs/2505.17150 | https://arxiv.org/pdf/2505.17150v1 | 2505.17150 | 10.14428/esann/2025.es2025-182 | 3 | 0 | false | null | The European Symposium on Artificial Neural Networks | 0.1684 |
b1bf0e90ff498848e4f7a5e70dd339e95900182d09a2a842edd8bc6bbaf39fdd | [
"arxiv",
"semantic_scholar"
] | Privacy-Aware Lifelong Learning | Lifelong learning algorithms enable models to incrementally acquire new knowledge without forgetting previously learned information. Contrarily, the field of machine unlearning focuses on explicitly forgetting certain previous knowledge from pretrained models when requested, in order to comply with data privacy regulat... | [
"Ozan Γzdenizci",
"Elmar Rueckert",
"Robert Legenstein"
] | [
"cs.LG"
] | [
"Computer Science"
] | 2025-05-16T00:00:00 | https://arxiv.org/abs/2505.10941 | https://arxiv.org/pdf/2505.10941v1 | 2505.10941 | 10.48550/arXiv.2505.10941 | 2 | 0 | false | null | International Conference on Learning Representations | 0.1616 |
7966ce1f169d52ae2a78c6512bbeff627fdac9ee7e467615e071e975cec9619d | [
"arxiv",
"semantic_scholar"
] | Preserving Plasticity in Continual Learning with Adaptive Linearity Injection | Loss of plasticity in deep neural networks is the gradual reduction in a model's capacity to incrementally learn and has been identified as a key obstacle to learning in non-stationary problem settings. Recent work has shown that deep linear networks tend to be resilient towards loss of plasticity. Motivated by this ob... | [
"Seyed Roozbeh Razavi Rohani",
"Khashayar Khajavi",
"Wesley Chung",
"Mo Chen",
"Sharan Vaswani"
] | [
"cs.LG",
"cs.AI"
] | [
"Computer Science"
] | 2025-05-14T00:00:00 | https://arxiv.org/abs/2505.09486 | https://arxiv.org/pdf/2505.09486v1 | 2505.09486 | 10.48550/arXiv.2505.09486 | 1 | 0 | false | null | arXiv.org | 0.1593 |
a83bd4b4abae83fd80c4431cc32f7b8cfe9634e91b4d3b6675921373527417d6 | [
"arxiv",
"semantic_scholar"
] | Bayesian continual learning and forgetting in neural networks | Biological synapses effortlessly balance memory retention and flexibility, yet artificial neural networks still struggle with the extremes of catastrophic forgetting and catastrophic remembering. Here, we introduce Metaplasticity from Synaptic Uncertainty (MESU), a Bayesian framework that updates network parameters acc... | [
"Djohan Bonnet",
"Kellian Cottart",
"Tifenn Hirtzlin",
"Tarcisius Januel",
"Thomas Dalgaty",
"Elisa Vianello",
"Damien Querlioz"
] | [
"cs.LG"
] | [
"Computer Science",
"Medicine"
] | 2025-04-18T00:00:00 | https://arxiv.org/abs/2504.13569 | https://arxiv.org/pdf/2504.13569v1 | 2504.13569 | 10.1038/s41467-025-64601-w | 13 | 2 | false | null | Nature Communications | 0.2865 |
4b11f0cafbf60bf0b47407c4103b255f978ad4c4b2cbff2967ce954f2898a7c5 | [
"arxiv",
"semantic_scholar"
] | Self-Controlled Dynamic Expansion Model for Continual Learning | Continual Learning (CL) epitomizes an advanced training paradigm wherein prior data samples remain inaccessible during the acquisition of new tasks. Numerous investigations have delved into leveraging a pre-trained Vision Transformer (ViT) to enhance model efficacy in continual learning. Nonetheless, these approaches t... | [
"Runqing Wu",
"Kaihui Huang",
"Hanyi Zhang",
"Fei Ye"
] | [
"cs.LG",
"cs.AI"
] | [
"Computer Science"
] | 2025-04-14T00:00:00 | https://arxiv.org/abs/2504.10561 | https://arxiv.org/pdf/2504.10561v2 | 2504.10561 | 10.48550/arXiv.2504.10561 | 1 | 0 | false | null | arXiv.org | 0.1249 |
5755c5c44f1c4a0b4222e1c0b4444ec4ea84969a2018bdab0d153b6d52f0af55 | [
"arxiv",
"semantic_scholar"
] | Learning and Improving Backgammon Strategy | A novel approach to learning is presented, combining features of on-line and off-line methods to achieve considerable performance in the task of learning a backgammon value function in a process that exploits the processing power of parallel supercomputers. The off-line methods comprise a set of techniques for parallel... | [
"Gregory R. Galperin"
] | [
"cs.LG",
"cs.AI",
"cs.NE"
] | [
"Computer Science"
] | 2025-04-03T00:00:00 | https://arxiv.org/abs/2504.02221 | https://arxiv.org/pdf/2504.02221v1 | 2504.02221 | 10.48550/arXiv.2504.02221 | 0 | 0 | false | null | arXiv.org | 0.1123 |
138ee58c411e657c87d7eb776522b248cdd3fce955424ba9786262a992db1390 | [
"arxiv",
"semantic_scholar"
] | Global Convergence of Continual Learning on Non-IID Data | Continual learning, which aims to learn multiple tasks sequentially, has gained extensive attention. However, most existing work focuses on empirical studies, and the theoretical aspect remains under-explored. Recently, a few investigations have considered the theory of continual learning only for linear regressions, e... | [
"Fei Zhu",
"Yujing Liu",
"Wenzhuo Liu",
"Zhaoxiang Zhang"
] | [
"cs.LG"
] | [
"Computer Science"
] | 2025-03-24T00:00:00 | https://arxiv.org/abs/2503.18511 | https://arxiv.org/pdf/2503.18511v1 | 2503.18511 | 10.48550/arXiv.2503.18511 | 3 | 0 | false | null | arXiv.org | 0.1505 |
4c037600884632076f47305df791ff299a9940eb335895e6610f4086cb17ac86 | [
"arxiv",
"semantic_scholar"
] | Birds look like cars: Adversarial analysis of intrinsically interpretable deep learning | A common belief is that intrinsically interpretable deep learning models ensure a correct, intuitive understanding of their behavior and offer greater robustness against accidental errors or intentional manipulation. However, these beliefs have not been comprehensively verified, and growing evidence casts doubt on them... | [
"Hubert Baniecki",
"Przemyslaw Biecek"
] | [
"cs.LG",
"cs.CV"
] | [
"Computer Science"
] | 2025-03-11T00:00:00 | https://arxiv.org/abs/2503.08636 | https://arxiv.org/pdf/2503.08636v2 | 2503.08636 | 10.1007/s10994-025-06896-w | 4 | 0 | false | null | Machine-mediated learning | 0.1747 |
6e639f60a69c906f36c3fc1cb2006e42b0f7f8396d1bb81f7e61c749c297fe3e | [
"arxiv",
"semantic_scholar"
] | A Good Start Matters: Enhancing Continual Learning with Data-Driven Weight Initialization | To adapt to real-world data streams, continual learning (CL) systems must rapidly learn new concepts while preserving and utilizing prior knowledge. When it comes to adding new information to continually-trained deep neural networks (DNNs), classifier weights for newly encountered categories are typically initialized r... | [
"Md Yousuf Harun",
"Christopher Kanan"
] | [
"cs.LG",
"cs.CV"
] | [
"Computer Science"
] | 2025-03-09T00:00:00 | https://arxiv.org/abs/2503.06385 | https://arxiv.org/pdf/2503.06385v2 | 2503.06385 | 10.48550/arXiv.2503.06385 | 2 | 0 | false | null | arXiv.org | 0.1193 |
5cf2b10c5bea7cb4854bd19497630db3841652cd47e003075ac7f05f6c0f0bac | [
"arxiv",
"semantic_scholar"
] | Beyond Cosine Decay: On the effectiveness of Infinite Learning Rate Schedule for Continual Pre-training | The ever-growing availability of unlabeled data presents both opportunities and challenges for training artificial intelligence systems. While self-supervised learning (SSL) has emerged as a powerful paradigm for extracting meaningful representations from vast amounts of unlabeled data, existing methods still struggle ... | [
"Vaibhav Singh",
"Paul Janson",
"Paria Mehrbod",
"Adam Ibrahim",
"Irina Rish",
"Eugene Belilovsky",
"Benjamin ThΓ©rien"
] | [
"cs.LG"
] | [
"Computer Science"
] | 2025-03-04T00:00:00 | https://arxiv.org/abs/2503.02844 | https://arxiv.org/pdf/2503.02844v3 | 2503.02844 | 10.48550/arXiv.2503.02844 | 5 | 0 | false | null | arXiv.org | 0.1945 |
a03ab28d3809148a2d2e39d6eebc131391e6343d26a67dc4ca51fe43c5c823c8 | [
"arxiv",
"semantic_scholar"
] | RIZE: Adaptive Regularization for Imitation Learning | We propose a novel Inverse Reinforcement Learning (IRL) method that mitigates the rigidity of fixed reward structures and the limited flexibility of implicit reward regularization. Building on the Maximum Entropy IRL framework, our approach incorporates a squared temporal-difference (TD) regularizer with adaptive targe... | [
"Adib Karimi",
"Mohammad Mehdi Ebadzadeh"
] | [
"cs.LG",
"cs.AI",
"cs.RO"
] | [
"Computer Science"
] | 2025-02-27T00:00:00 | https://arxiv.org/abs/2502.20089 | https://arxiv.org/pdf/2502.20089v3 | 2502.20089 | null | 1 | 1 | true | https://github.com/adibka/RIZE | Transactions on Machine Learning Research (11/2025) | 0.1505 |
a8aef3298d369ec3cd84eff5800403c74479210f739bf13e7c2e8cd0e7b8b09a | [
"arxiv",
"semantic_scholar"
] | Eidetic Learning: an Efficient and Provable Solution to Catastrophic Forgetting | Catastrophic forgetting -- the phenomenon of a neural network learning a task t1 and losing the ability to perform it after being trained on some other task t2 -- is a long-standing problem for neural networks [McCloskey and Cohen, 1989]. We present a method, Eidetic Learning, that provably solves catastrophic forgetti... | [
"Nicholas Dronen",
"Randall Balestriero"
] | [
"cs.LG"
] | [
"Computer Science"
] | 2025-02-13T00:00:00 | https://arxiv.org/abs/2502.09500 | https://arxiv.org/pdf/2502.09500v2 | 2502.09500 | 10.48550/arXiv.2502.09500 | 0 | 0 | true | https://github.com/amazon-science/eideticnet-training | arXiv.org | 0.0868 |
fba79a3a78c214d7b9961510c4b0acb386eddd122b7499bebb21ac21245915a1 | [
"arxiv",
"semantic_scholar"
] | Predicting concentration levels of air pollutants by transfer learning and recurrent neural network | Air pollution (AP) poses a great threat to human health, and people are paying more attention than ever to its prediction. Accurate prediction of AP helps people to plan for their outdoor activities and aids protecting human health. In this paper, long-short term memory (LSTM) recurrent neural networks (RNNs) have been... | [
"Iat Hang Fong",
"Tengyue Li",
"Simon Fong",
"Raymond K. Wong",
"Antonio J. TallΓ³n-Ballesteros"
] | [
"cs.LG",
"cs.NE",
"physics.ao-ph"
] | [
"Computer Science",
"Physics"
] | 2025-01-30T00:00:00 | https://arxiv.org/abs/2502.01654 | https://arxiv.org/pdf/2502.01654v1 | 2502.01654 | 10.1016/j.knosys.2020.105622 | 85 | 3 | false | null | Knowledge-Based Systems | 0.4836 |
6ae19ac747da297caf8b5c7c3c1bed651226e25e3f65c6b747d7c749c7b6cfe5 | [
"arxiv",
"semantic_scholar"
] | U-Fair: Uncertainty-based Multimodal Multitask Learning for Fairer Depression Detection | Machine learning bias in mental health is becoming an increasingly pertinent challenge. Despite promising efforts indicating that multitask approaches often work better than unitask approaches, there is minimal work investigating the impact of multitask learning on performance and fairness in depression detection nor l... | [
"Jiaee Cheong",
"Aditya Bangar",
"Sinan Kalkan",
"Hatice Gunes"
] | [
"cs.LG"
] | [
"Computer Science"
] | 2025-01-16T00:00:00 | https://arxiv.org/abs/2501.09687 | https://arxiv.org/pdf/2501.09687v1 | 2501.09687 | 10.48550/arXiv.2501.09687 | 17 | 1 | false | null | Proceedings of Machine Learning Research 2024 | 0.3138 |
fef0b3a2b9108b59024bf79ee189ecfdb9c6cb73f4b3356b5c1e05a268bd4cbf | [
"arxiv",
"semantic_scholar"
] | An Empirical Analysis of Federated Learning Models Subject to Label-Flipping Adversarial Attack | In this paper, we empirically analyze adversarial attacks on selected federated learning models. The specific learning models considered are Multinominal Logistic Regression (MLR), Support Vector Classifier (SVC), Multilayer Perceptron (MLP), Convolution Neural Network (CNN), %Recurrent Neural Network (RNN), Random For... | [
"Kunal Bhatnagar",
"Sagana Chattanathan",
"Angela Dang",
"Bhargav Eranki",
"Ronnit Rana",
"Charan Sridhar",
"Siddharth Vedam",
"Angie Yao",
"Mark Stamp"
] | [
"cs.LG"
] | [
"Computer Science"
] | 2024-12-24T00:00:00 | https://arxiv.org/abs/2412.18507 | https://arxiv.org/pdf/2412.18507v1 | 2412.18507 | 10.48550/arXiv.2412.18507 | 2 | 0 | false | null | arXiv.org | 0.1193 |
ed0cb6cc6cc2f05df477ba7fbcf4871bfd3c3885776ebcc339cb35d20d0f4e75 | [
"arxiv",
"semantic_scholar"
] | Modality-Inconsistent Continual Learning of Multimodal Large Language Models | In this paper, we introduce Modality-Inconsistent Continual Learning (MICL), a new continual learning scenario for Multimodal Large Language Models (MLLMs) that involves tasks with inconsistent modalities (image, audio, or video) and varying task types (captioning or question-answering). Unlike existing vision-only or ... | [
"Weiguo Pian",
"Shijian Deng",
"Shentong Mo",
"Mingrui Liu",
"Yunhui Guo",
"Yapeng Tian"
] | [
"cs.LG",
"cs.AI",
"cs.CL",
"cs.CV",
"cs.SD",
"eess.AS"
] | [
"Computer Science",
"Engineering"
] | 2024-12-17T00:00:00 | https://arxiv.org/abs/2412.13050 | https://arxiv.org/pdf/2412.13050v2 | 2412.13050 | 10.48550/arXiv.2412.13050 | 6 | 0 | false | null | arXiv.org | 0.2113 |
6650f5ee057d450ca3b997ca6c61d6f5c089ce00521711731b7edbc668eca90c | [
"arxiv",
"semantic_scholar"
] | Learning to Navigate in Mazes with Novel Layouts using Abstract Top-down Maps | Learning navigation capabilities in different environments has long been one of the major challenges in decision-making. In this work, we focus on zero-shot navigation ability using given abstract $2$-D top-down maps. Like human navigation by reading a paper map, the agent reads the map as an image when navigating in a... | [
"Linfeng Zhao",
"Lawson L. S. Wong"
] | [
"cs.LG",
"cs.AI",
"cs.RO"
] | [
"Computer Science"
] | 2024-12-16T00:00:00 | https://arxiv.org/abs/2412.12024 | https://arxiv.org/pdf/2412.12024v1 | 2412.12024 | 10.48550/arXiv.2412.12024 | 3 | 0 | false | null | Journal-ref: Reinforcement Learning Journal, Volume 5, 2024, Pages 2359-2372 | 0.1505 |
9a851e99343d8fa3d43b96b96a7f53f146ebdd64cafe9654a312dcb86d14de6b | [
"arxiv",
"semantic_scholar"
] | GLL: A Differentiable Graph Learning Layer for Neural Networks | Standard deep learning architectures used for classification generate label predictions with a projection head and softmax activation function. Although successful, these methods fail to leverage the relational information between samples for generating label predictions. In recent works, graph-based learning technique... | [
"Jason Brown",
"Bohan Chen",
"Harris Hardiman-Mostow",
"Jeff Calder",
"Andrea L. Bertozzi"
] | [
"cs.LG",
"stat.ML"
] | [
"Computer Science",
"Mathematics"
] | 2024-12-11T00:00:00 | https://arxiv.org/abs/2412.08016 | https://arxiv.org/pdf/2412.08016v2 | 2412.08016 | 10.48550/arXiv.2412.08016 | 0 | 0 | false | null | arXiv.org | 0 |
336f47d06538964e249f6ef1ec40751fb4756b8a10a5d5cb86c846eb9e9b35ea | [
"arxiv",
"semantic_scholar"
] | Towards Fast Safe Online Reinforcement Learning via Policy Finetuning | The high costs and risks involved in extensive environment interactions hinder the practical application of current online safe reinforcement learning (RL) methods. While offline safe RL addresses this by learning policies from static datasets, the performance therein is usually limited due to reliance on data quality ... | [
"Keru Chen",
"Honghao Wei",
"Zhigang Deng",
"Sen Lin"
] | [
"cs.LG",
"cs.AI"
] | [
"Computer Science"
] | 2024-12-05T00:00:00 | https://arxiv.org/abs/2412.04426 | https://arxiv.org/pdf/2412.04426v4 | 2412.04426 | null | 5 | 0 | false | null | Transactions on Machine Learning Research (TMLR), 2026 | 0.1945 |
d1f352e14340829e04ef51162c4fa6a002b7fd0366724f7f86df8884bea95043 | [
"arxiv",
"semantic_scholar"
] | Robust Offline Reinforcement Learning with Linearly Structured f-Divergence Regularization | The Robust Regularized Markov Decision Process (RRMDP) is proposed to learn policies robust to dynamics shifts by adding regularization to the transition dynamics in the value function. Existing methods mostly use unstructured regularization, potentially leading to conservative policies under unrealistic transitions. T... | [
"Cheng Tang",
"Zhishuai Liu",
"Pan Xu"
] | [
"cs.LG",
"cs.AI",
"cs.RO",
"stat.ML"
] | [
"Computer Science",
"Mathematics"
] | 2024-11-27T00:00:00 | https://arxiv.org/abs/2411.18612 | https://arxiv.org/pdf/2411.18612v2 | 2411.18612 | 10.48550/arXiv.2411.18612 | 6 | 0 | false | null | International Conference on Machine Learning | 0.2113 |
c3fcd4e5924e106b357a800e9defcc8bd682502f41c3f5c502c2564159630e5e | [
"arxiv",
"semantic_scholar"
] | Learning Explainable Treatment Policies with Clinician-Informed Representations: A Practical Approach | Digital health interventions (DHIs) and remote patient monitoring (RPM) have shown great potential in improving chronic disease management through personalized care. However, barriers like limited efficacy and workload concerns hinder adoption of existing DHIs; while limited sample sizes and lack of interpretability li... | [
"Johannes O. Ferstad",
"Emily B. Fox",
"David Scheinker",
"Ramesh Johari"
] | [
"cs.LG",
"cs.AI",
"stat.AP",
"stat.ML"
] | [
"Computer Science",
"Mathematics"
] | 2024-11-26T00:00:00 | https://arxiv.org/abs/2411.17570 | https://arxiv.org/pdf/2411.17570v1 | 2411.17570 | 10.48550/arXiv.2411.17570 | 1 | 0 | true | https://github.com/jferstad/ml4h-explainable-policies | Proceedings of the 4th Machine Learning for Health Symposium, PMLR 259:325-349, 2025 | 0.0753 |
80bc76755004ee91762cff2835acf7bea5bdf12d6a1897b4206fb80a1e79e63d | [
"arxiv",
"semantic_scholar"
] | Slowing Down Forgetting in Continual Learning | A common challenge in continual learning (CL) is catastrophic forgetting, where the performance on old tasks drops after new, additional tasks are learned. In this paper, we propose a novel framework called ReCL to slow down forgetting in CL. Our framework exploits an implicit bias of gradient-based neural networks due... | [
"Pascal Janetzky",
"Tobias Schlagenhauf",
"Stefan Feuerriegel"
] | [
"cs.LG",
"cs.AI",
"cs.CV"
] | [
"Computer Science"
] | 2024-11-11T00:00:00 | https://arxiv.org/abs/2411.06916 | https://arxiv.org/pdf/2411.06916v2 | 2411.06916 | 10.48550/arXiv.2411.06916 | 0 | 0 | false | null | arXiv.org | 0 |
52ad4e434a517294fe71e828e65bfb28d5226553ddfdd3154819f2e6cebea43f | [
"arxiv",
"semantic_scholar"
] | Return Augmented Decision Transformer for Off-Dynamics Reinforcement Learning | We study offline off-dynamics reinforcement learning (RL) to utilize data from an easily accessible source domain to enhance policy learning in a target domain with limited data. Our approach centers on return-conditioned supervised learning (RCSL), particularly focusing on Decision Transformer (DT) type frameworks, wh... | [
"Ruhan Wang",
"Yu Yang",
"Zhishuai Liu",
"Dongruo Zhou",
"Pan Xu"
] | [
"cs.LG",
"cs.AI",
"cs.RO",
"stat.ML"
] | [
"Computer Science",
"Mathematics"
] | 2024-10-30T00:00:00 | https://arxiv.org/abs/2410.23450 | https://arxiv.org/pdf/2410.23450v2 | 2410.23450 | 10.48550/arXiv.2410.23450 | 14 | 0 | false | null | Transactions on Machine Learning Research, 2026 | 0.294 |
d244f818afce974bf87dbe7d9fffb5c86a95b0950adbd27675118e23aea1a544 | [
"arxiv",
"semantic_scholar"
] | The Effects of Multi-Task Learning on ReLU Neural Network Functions | This paper studies the properties of solutions to multi-task shallow ReLU neural network learning problems, wherein the network is trained to fit a dataset with minimal sum of squared weights. Remarkably, the solutions learned for each individual task resemble those obtained by solving a kernel regression problem, reve... | [
"Julia Nakhleh",
"Joseph Shenouda",
"Robert D. Nowak"
] | [
"stat.ML",
"cs.LG"
] | [
"Computer Science",
"Mathematics"
] | 2024-10-29T00:00:00 | https://arxiv.org/abs/2410.21696 | https://arxiv.org/pdf/2410.21696v4 | 2410.21696 | 10.48550/arXiv.2410.21696 | 1 | 0 | false | null | arXiv.org | 0.0753 |
8012bfeca2d752a9b972c82020ce91042e5aacca1df4a4e23e40402d53bb8028 | [
"arxiv",
"semantic_scholar"
] | Improving Multimodal Large Language Models Using Continual Learning | Generative large language models (LLMs) exhibit impressive capabilities, which can be further augmented by integrating a pre-trained vision model into the original LLM to create a multimodal LLM (MLLM). However, this integration often significantly decreases performance on natural language understanding and generation ... | [
"Shikhar Srivastava",
"Md Yousuf Harun",
"Robik Shrestha",
"Christopher Kanan"
] | [
"cs.CL",
"cs.CV",
"cs.LG"
] | [
"Computer Science"
] | 2024-10-25T00:00:00 | https://arxiv.org/abs/2410.19925 | https://arxiv.org/pdf/2410.19925v2 | 2410.19925 | 10.48550/arXiv.2410.19925 | 4 | 0 | false | null | arXiv.org | 0.1747 |
2043f6e7b9a8be39d290778da6b9ea8dec44b9228f20671bb63ea5788455a711 | [
"arxiv",
"semantic_scholar"
] | SNAP: Stopping Catastrophic Forgetting in Hebbian Learning with Sigmoidal Neuronal Adaptive Plasticity | Artificial Neural Networks (ANNs) suffer from catastrophic forgetting, where the learning of new tasks causes the catastrophic forgetting of old tasks. Existing Machine Learning (ML) algorithms, including those using Stochastic Gradient Descent (SGD) and Hebbian Learning typically update their weights linearly with exp... | [
"Tianyi Xu",
"Patrick Zheng",
"Shiyan Liu",
"Sicheng Lyu",
"Isabeau PrΓ©mont-Schwarz"
] | [
"cs.NE",
"cs.AI",
"cs.LG"
] | [
"Computer Science"
] | 2024-10-20T00:00:00 | https://arxiv.org/abs/2410.15318 | https://arxiv.org/pdf/2410.15318v1 | 2410.15318 | 10.48550/arXiv.2410.15318 | 0 | 0 | false | null | arXiv.org | 0 |
7e9282a7a056712a6f1547fc1a5959dab7ad1eb48d0391e99692c3534698b2eb | [
"arxiv",
"semantic_scholar"
] | Continual Deep Reinforcement Learning to Prevent Catastrophic Forgetting in Jamming Mitigation | Deep Reinforcement Learning (DRL) has been highly effective in learning from and adapting to RF environments and thus detecting and mitigating jamming effects to facilitate reliable wireless communications. However, traditional DRL methods are susceptible to catastrophic forgetting (namely forgetting old tasks when lea... | [
"Kemal Davaslioglu",
"Sastry Kompella",
"Tugba Erpek",
"Yalin E. Sagduyu"
] | [
"cs.LG",
"cs.AI",
"cs.NI"
] | [
"Computer Science"
] | 2024-10-14T00:00:00 | https://arxiv.org/abs/2410.10521 | https://arxiv.org/pdf/2410.10521v1 | 2410.10521 | 10.1109/MILCOM61039.2024.10773861 | 7 | 0 | false | null | IEEE Military Communications Conference | 0.2258 |
d6200f974bff103076f9972960728197bbf44cdf6d5796bf7027abcd8b8d0570 | [
"arxiv",
"semantic_scholar"
] | Metalic: Meta-Learning In-Context with Protein Language Models | Predicting the biophysical and functional properties of proteins is essential for in silico protein design. Machine learning has emerged as a promising technique for such prediction tasks. However, the relative scarcity of in vitro annotations means that these models often have little, or no, specific data on the desir... | [
"Jacob Beck",
"Shikha Surana",
"Manus McAuliffe",
"Oliver Bent",
"Thomas D. Barrett",
"Juan Jose Garau Luis",
"Paul Duckworth"
] | [
"cs.LG"
] | [
"Computer Science"
] | 2024-10-10T00:00:00 | https://arxiv.org/abs/2410.08355 | https://arxiv.org/pdf/2410.08355v3 | 2410.08355 | 10.48550/arXiv.2410.08355 | 4 | 0 | true | https://github.com/instadeepai/metalic | International Conference on Learning Representations | 0.1747 |
79659c2c098850941dda1357e7297579a3dd8988bcc48a7df745726018b67441 | [
"arxiv",
"semantic_scholar"
] | Scalable Mechanistic Neural Networks for Differential Equations and Machine Learning | We propose Scalable Mechanistic Neural Network (S-MNN), an enhanced neural network framework designed for scientific machine learning applications involving long temporal sequences. By reformulating the original Mechanistic Neural Network (MNN) (Pervez et al., 2024), we reduce the computational time and space complexit... | [
"Jiale Chen",
"Dingling Yao",
"Adeel Pervez",
"Dan Alistarh",
"Francesco Locatello"
] | [
"cs.LG",
"math.NA"
] | [
"Computer Science",
"Mathematics"
] | 2024-10-08T00:00:00 | https://arxiv.org/abs/2410.06074 | https://arxiv.org/pdf/2410.06074v3 | 2410.06074 | 10.48550/arXiv.2410.06074 | 4 | 0 | true | https://github.com/IST-DASLab/ScalableMNN | International Conference on Learning Representations | 0.1747 |
279d0bc6be060ae753ae9adf8f74d3f12a2fe4c2df3fb9430845fd7e5e761dbb | [
"arxiv",
"semantic_scholar"
] | Recent Advances of Multimodal Continual Learning: A Comprehensive Survey | Continual learning (CL) aims to empower machine learning models to learn continually from new data, while building upon previously acquired knowledge without forgetting. As models have evolved from small to large pre-trained architectures, and from supporting unimodal to multimodal data, multimodal continual learning (... | [
"Dianzhi Yu",
"Xinni Zhang",
"Yankai Chen",
"Aiwei Liu",
"Yifei Zhang",
"Philip S. Yu",
"Irwin King"
] | [
"cs.LG",
"cs.AI"
] | [
"Medicine",
"Computer Science"
] | 2024-10-07T00:00:00 | https://arxiv.org/abs/2410.05352 | https://arxiv.org/pdf/2410.05352v3 | 2410.05352 | 10.48550/arXiv.2410.05352 | 45 | 1 | true | https://github.com/LucyDYu/Awesome-Multimodal-Continual-Learning | IEEE Transactions on Neural Networks and Learning Systems | 0.4157 |
03a355a6a4ec6179d93ce97995cc020129ae233fbc6cdec422c34d34a3236d7b | [
"arxiv",
"semantic_scholar"
] | Optimal Protocols for Continual Learning via Statistical Physics and Control Theory | Artificial neural networks often struggle with catastrophic forgetting when learning multiple tasks sequentially, as training on new tasks degrades the performance on previously learned tasks. Recent theoretical work has addressed this issue by analysing learning curves in synthetic frameworks under predefined training... | [
"Francesco Mori",
"Stefano Sarao Mannelli",
"Francesca Mignacco"
] | [
"cs.LG",
"cond-mat.dis-nn",
"cond-mat.stat-mech"
] | [
"Computer Science",
"Physics"
] | 2024-09-26T00:00:00 | https://arxiv.org/abs/2409.18061 | https://arxiv.org/pdf/2409.18061v3 | 2409.18061 | 10.1088/1742-5468/adf296 | 15 | 1 | false | null | International Conference on Learning Representations | 0.301 |
eabdeafb5e974afa235ed22714c1ad2ee61cb6dd60a8afaf6f5622d4d2165997 | [
"arxiv",
"semantic_scholar"
] | Patch-Based Contrastive Learning and Memory Consolidation for Online Unsupervised Continual Learning | We focus on a relatively unexplored learning paradigm known as {\em Online Unsupervised Continual Learning} (O-UCL), where an agent receives a non-stationary, unlabeled data stream and progressively learns to identify an increasing number of classes. This paradigm is designed to model real-world applications where enco... | [
"Cameron Taylor",
"Vassilis Vassiliades",
"Constantine Dovrolis"
] | [
"cs.LG",
"cs.CV"
] | [
"Computer Science"
] | 2024-09-24T00:00:00 | https://arxiv.org/abs/2409.16391 | https://arxiv.org/pdf/2409.16391v1 | 2409.16391 | 10.48550/arXiv.2409.16391 | 1 | 0 | false | null | null | 0.0753 |
1e352c84ec288d33aebe43af05b954e283899ecfd570f19d67607b8c3a41c1da | [
"arxiv",
"semantic_scholar"
] | A Contrastive Symmetric Forward-Forward Algorithm (SFFA) for Continual Learning Tasks | The so-called Forward-Forward Algorithm (FFA) has recently gained momentum as an alternative to the conventional back-propagation algorithm for neural network learning, yielding competitive performance across various modeling tasks. By replacing the backward pass of gradient back-propagation with two contrastive forwar... | [
"Erik B. Terres-Escudero",
"Javier Del Ser",
"Pablo Garcia Bringas"
] | [
"cs.LG"
] | [
"Computer Science"
] | 2024-09-11T00:00:00 | https://arxiv.org/abs/2409.07387 | https://arxiv.org/pdf/2409.07387v2 | 2409.07387 | 10.48550/arXiv.2409.07387 | 2 | 0 | false | null | null | 0.1193 |
da3176728c186e3d4604f510c332aac0f53dfb431b383d93ea9a0b9e76192fd0 | [
"arxiv",
"semantic_scholar"
] | Buffer-based Gradient Projection for Continual Federated Learning | Continual Federated Learning (CFL) is essential for enabling real-world applications where multiple decentralized clients adaptively learn from continuous data streams. A significant challenge in CFL is mitigating catastrophic forgetting, where models lose previously acquired knowledge when learning new information. Ex... | [
"Shenghong Dai",
"Jy-yong Sohn",
"Yicong Chen",
"S M Iftekharul Alam",
"Ravikumar Balakrishnan",
"Suman Banerjee",
"Nageen Himayat",
"Kangwook Lee"
] | [
"cs.LG",
"cs.DC"
] | [
"Computer Science"
] | 2024-09-03T00:00:00 | https://arxiv.org/abs/2409.01585 | https://arxiv.org/pdf/2409.01585v1 | 2409.01585 | 10.48550/arXiv.2409.01585 | 4 | 0 | true | https://github.com/shenghongdai/Fed-A-GEM | null | 0.1747 |
f5f33f268f5c99164da5c472f7b0e369b5e4f69b344d5dbf79418d4ca21e2902 | [
"arxiv",
"semantic_scholar"
] | Continual learning with the neural tangent ensemble | A natural strategy for continual learning is to weigh a Bayesian ensemble of fixed functions. This suggests that if a (single) neural network could be interpreted as an ensemble, one could design effective algorithms that learn without forgetting. To realize this possibility, we observe that a neural network classifier... | [
"Ari S. Benjamin",
"Christian Pehle",
"Kyle Daruwalla"
] | [
"cs.LG",
"cs.NE"
] | [
"Computer Science"
] | 2024-08-30T00:00:00 | https://arxiv.org/abs/2408.17394 | https://arxiv.org/pdf/2408.17394v2 | 2408.17394 | 10.48550/arXiv.2408.17394 | 3 | 0 | false | null | Neural Information Processing Systems | 0.1505 |
86e72946f2dd7bfa094f7329d135f0c4be66cbb44ff1346f445d8d8ab2d6ac66 | [
"arxiv",
"semantic_scholar"
] | Learning Multi-Index Models with Neural Networks via Mean-Field Langevin Dynamics | We study the problem of learning multi-index models in high-dimensions using a two-layer neural network trained with the mean-field Langevin algorithm. Under mild distributional assumptions on the data, we characterize the effective dimension $d_{\mathrm{eff}}$ that controls both sample and computational complexity by ... | [
"Alireza Mousavi-Hosseini",
"Denny Wu",
"Murat A. Erdogdu"
] | [
"stat.ML",
"cs.LG"
] | [
"Computer Science",
"Mathematics"
] | 2024-08-14T00:00:00 | https://arxiv.org/abs/2408.07254 | https://arxiv.org/pdf/2408.07254v2 | 2408.07254 | 10.48550/arXiv.2408.07254 | 13 | 0 | false | null | International Conference on Learning Representations | 0.2865 |
3f5265d68735dc4ec97a6e975ef93031500d986c5cacf4de40bbac2cb0a7cdf4 | [
"arxiv",
"semantic_scholar"
] | Learning to Learn without Forgetting using Attention | Continual learning (CL) refers to the ability to continually learn over time by accommodating new knowledge while retaining previously learned experience. While this concept is inherent in human learning, current machine learning methods are highly prone to overwrite previously learned patterns and thus forget past exp... | [
"Anna Vettoruzzo",
"Joaquin Vanschoren",
"Mohamed-Rafik Bouguelia",
"Thorsteinn RΓΆgnvaldsson"
] | [
"cs.LG",
"cs.CV"
] | [
"Computer Science"
] | 2024-08-06T00:00:00 | https://arxiv.org/abs/2408.03219 | https://arxiv.org/pdf/2408.03219v2 | 2408.03219 | 10.48550/arXiv.2408.03219 | 2 | 0 | false | null | null | 0.1193 |
cd080e33193dc5b9ca8a3e10ffb3580a10a8712527b3283c6c07e74b4d8a81bd | [
"arxiv",
"semantic_scholar"
] | Diffusion Augmented Agents: A Framework for Efficient Exploration and Transfer Learning | We introduce Diffusion Augmented Agents (DAAG), a novel framework that leverages large language models, vision language models, and diffusion models to improve sample efficiency and transfer learning in reinforcement learning for embodied agents. DAAG hindsight relabels the agent's past experience by using diffusion mo... | [
"Norman Di Palo",
"Leonard Hasenclever",
"Jan Humplik",
"Arunkumar Byravan"
] | [
"cs.LG",
"cs.AI",
"cs.RO"
] | [
"Computer Science"
] | 2024-07-30T00:00:00 | https://arxiv.org/abs/2407.20798 | https://arxiv.org/pdf/2407.20798v1 | 2407.20798 | 10.48550/arXiv.2407.20798 | 5 | 0 | false | null | null | 0.1945 |
63a8a658e4435a28415dacc4b78d1091e74508f12a3b348539cb583cc6d788c7 | [
"arxiv",
"semantic_scholar"
] | Gradient Boosting Reinforcement Learning | We present Gradient Boosting Reinforcement Learning (GBRL), a framework that adapts the strengths of gradient boosting trees (GBT) to reinforcement learning (RL) tasks. While neural networks (NNs) have become the de facto choice for RL, they face significant challenges with structured and categorical features and tend ... | [
"Benjamin Fuhrer",
"Chen Tessler",
"Gal Dalal"
] | [
"cs.LG",
"cs.AI"
] | [
"Computer Science"
] | 2024-07-11T00:00:00 | https://arxiv.org/abs/2407.08250 | https://arxiv.org/pdf/2407.08250v2 | 2407.08250 | 10.48550/arXiv.2407.08250 | 5 | 0 | false | null | International Conference on Machine Learning | 0.1945 |
eea226d772d1c320a9074c69e411018fb2842a97de61e6f1e71a266d85b333ff | [
"arxiv",
"semantic_scholar"
] | How to Leverage Predictive Uncertainty Estimates for Reducing Catastrophic Forgetting in Online Continual Learning | Many real-world applications require machine-learning models to be able to deal with non-stationary data distributions and thus learn autonomously over an extended period of time, often in an online setting. One of the main challenges in this scenario is the so-called catastrophic forgetting (CF) for which the learning... | [
"Giuseppe Serra",
"Ben Werner",
"Florian Buettner"
] | [
"cs.LG",
"cs.AI"
] | [
"Computer Science"
] | 2024-07-10T00:00:00 | https://arxiv.org/abs/2407.07668 | https://arxiv.org/pdf/2407.07668v3 | 2407.07668 | 10.48550/arXiv.2407.07668 | 7 | 0 | false | null | null | 0.2258 |
c9b3bdf3aa7df4821a997ff6b58f9263de964e63456938063a6e4c5bb5706250 | [
"arxiv",
"semantic_scholar"
] | The impact of model size on catastrophic forgetting in Online Continual Learning | This study investigates the impact of model size on Online Continual Learning performance, with a focus on catastrophic forgetting. Employing ResNet architectures of varying sizes, the research examines how network depth and width affect model performance in class-incremental learning using the SplitCIFAR-10 dataset. K... | [
"Eunhae Lee"
] | [
"cs.LG",
"cs.CV"
] | [
"Computer Science"
] | 2024-06-28T00:00:00 | https://arxiv.org/abs/2407.00176 | https://arxiv.org/pdf/2407.00176v1 | 2407.00176 | 10.48550/arXiv.2407.00176 | 3 | 0 | false | null | arXiv.org | 0.1505 |
c2158d623790a66666a21e9630c134ba184756d90dba986b2bdcae7300285dd0 | [
"arxiv",
"semantic_scholar"
] | Retrospective Feature Estimation for Continual Learning | The intrinsic capability to continuously learn a changing data stream is a desideratum of deep neural networks (DNNs). However, current DNNs suffer from catastrophic forgetting, which interferes with remembering past knowledge. To mitigate this issue, existing Continual Learning (CL) approaches often retain exemplars f... | [
"Nghia D. Nguyen",
"Hieu Trung Nguyen",
"Ang Li",
"Hoang Pham",
"Viet Anh Nguyen",
"Khoa D. Doan"
] | [
"cs.LG",
"cs.CV"
] | [
"Computer Science"
] | 2024-06-25T00:00:00 | https://arxiv.org/abs/2406.17381 | https://arxiv.org/pdf/2406.17381v2 | 2406.17381 | null | 0 | 0 | true | https://github.com/mail-research/retrospective-feature-estimation | null | 0 |
c49f55f2262da9c4078d9ae7f69474e34804c669fba2e9771ccc0c0746a82271 | [
"arxiv",
"semantic_scholar"
] | Learning Temporal Distances: Contrastive Successor Features Can Provide a Metric Structure for Decision-Making | Temporal distances lie at the heart of many algorithms for planning, control, and reinforcement learning that involve reaching goals, allowing one to estimate the transit time between two states. However, prior attempts to define such temporal distances in stochastic settings have been stymied by an important limitatio... | [
"Vivek Myers",
"Chongyi Zheng",
"Anca Dragan",
"Sergey Levine",
"Benjamin Eysenbach"
] | [
"cs.LG",
"cs.AI"
] | [
"Computer Science"
] | 2024-06-24T00:00:00 | https://arxiv.org/abs/2406.17098 | https://arxiv.org/pdf/2406.17098v2 | 2406.17098 | 10.48550/arXiv.2406.17098 | 48 | 4 | false | null | International Conference on Machine Learning | 0.4225 |
b7202610fc64bca15aeebf80c4d3b217ba154ed2dbfb06cf512d711320975f98 | [
"arxiv",
"semantic_scholar"
] | Towards evolution of Deep Neural Networks through contrastive Self-Supervised learning | Deep Neural Networks (DNNs) have been successfully applied to a wide range of problems. However, two main limitations are commonly pointed out. The first one is that they require long time to design. The other is that they heavily rely on labelled data, which can sometimes be costly and hard to obtain. In order to addr... | [
"Adriano Vinhas",
"JoΓ£o Correia",
"Penousal Machado"
] | [
"cs.NE",
"cs.AI",
"cs.LG"
] | [
"Computer Science"
] | 2024-06-20T00:00:00 | https://arxiv.org/abs/2406.14525 | https://arxiv.org/pdf/2406.14525v1 | 2406.14525 | 10.1109/CEC60901.2024.10611910 | 0 | 0 | false | null | IEEE Congress on Evolutionary Computation | 0 |
End of preview. Expand in Data Studio
Continual Learning Papers β FineSet
A research-paper dataset on Continual Learning Papers, assembled, deduplicated, and quality-scored by FineSet from arXiv and Semantic Scholar.
πΈ This is a dated snapshot β generated 2026-06-19. It is not auto-updated. Research on Continual Learning Papers moves fast β new papers land on arXiv every week. Want this same dataset refreshed daily, on a topic you choose? See the bottom. β
Why this dataset
- Quality-scored:
quality_scorefloat (0β1), blends citations with recency + code/venue signals β filter out the noise - Papers with code: 68 flagged via
has_codeβ find reproducible work fast - Deduplicated: arXiv + Semantic Scholar cross-referenced, duplicate records merged
- Clean JSONL: 431 records, one per line, normalized fields β no encoding garbage
Dataset details
- Records: 431
- Date range: 2019β2026
- Snapshot date: 2026-06-19 (frozen β see note above)
- Sources: arXiv, Semantic Scholar (cross-referenced, duplicates merged)
- arXiv categories: cs.LG
- Quality scoring: citations + recency + code/venue blend, 0β1 (p50=0.325, p90=0.557)
- Format: JSONL, one record per line
Fields
| Field | Type | Description |
|---|---|---|
| id | string | Deterministic SHA256 record id |
| sources | list | Which sources contributed (arxiv, semantic_scholar) |
| title | string | Paper title |
| abstract | string | Full abstract |
| authors | list | Author names |
| categories | list | arXiv category codes |
| fields_of_study | list | Semantic Scholar field tags |
| published_date | string | ISO 8601 date |
| url | string | arXiv abstract URL |
| pdf_url | string|null | Open-access PDF if available |
| arxiv_id | string|null | arXiv identifier |
| doi | string|null | DOI if available |
| citation_count | int | Citation count (Semantic Scholar) |
| influential_citation_count | int | Influential citations (Semantic Scholar) |
| has_code | bool | Code repo detected in the arXiv comment |
| code_url | string|null | GitHub URL if detected |
| venue | string|null | Publication venue |
| quality_score | float | 0β1, blended (citations + recency + code/venue) |
Quality score methodology
quality_score = max(impact, freshness), clamped to [0, 1], where:
- impact =
max( log10(citations+1)/4 , log10(influential_citations+1)/2 )β realized impact (0.5 at 100 citations, ~0.75 at 1,000, 1.0 at 10,000+). - freshness =
recency Γ (0.35 + 0.30Β·has_code + 0.20Β·has_venue)β a baseline for recent papers (so a strong paper published this week isn't scored 0 just for lacking citations), whererecencyis 1.0 for papers β€60 days old and decays linearly to 0 by ~18 months.
Old highly-cited papers score on impact; brand-new papers score on freshness; old uncited papers score ~0. Useful for filtering training data by quality, not just age.
π Want this on YOUR topic, updated daily?
This snapshot is frozen at 2026-06-19. The live FineSet pipeline keeps a dataset like this refreshed every day on whatever topic you describe β new papers in, dedup and quality scoring automatic, export as JSONL/Parquet or push straight to the Hub.
Tell me the topic you'd want and I'll run the pipeline on it β open a discussion on this dataset, it's free and it's how I decide what to build next.
β fineset.io β describe what you want to train on, get a dataset. Early-access waitlist open (referral skip available).
- Downloads last month
- 31