#continual learning
28 papers match
DECODE: Tackling Representation and Decision Degradation in Continual AI-Generated Image Detection
Zihao Cai, Xinghan Li, Ruiyan Yang +3
The paper introduces DECODE, a framework that addresses both representation and decision-level forgetting in continual learning for AI-generated image detection, using subspace div…
FedOGL: Combating Catastrophic Forgetting in Federated Open-World Multimodal Graph Learning
Zekai Chen, Haodong Lu, Shihao Li +5
The paper introduces FedOGL, a framework for federated multimodal graph learning that mitigates catastrophic forgetting by preserving semantic and structural memory through client-…
Post-Training at the Edge of Detectability: A Game-Theoretic Approach to Fine-Tuning
Keegan Harris, Brian W. Lee, Ian Waudby-Smith +3
The paper introduces a game‑theoretic framework for RL fine‑tuning that determines the KL regularization coefficient by treating the trade‑off between reward and deviation from a r…
Progressive Multimodal Alignment for Continual Instruction Tuning
Duzhen Zhang, Yahan Yu, Qiaoyi Su +2
The paper proposes Progressive Multimodal Alignment (PMA), a framework that adds expandable expert projectors and a routing mechanism to continually adapt visual-language alignment…
ForgetBench: Benchmarking Forgetting Dynamics of Long-Term Parametric Memory in Language Models
Ruxi Gu, Zhenliang Zhang, Wei Wang
The paper introduces ForgetBench, a benchmark for measuring how large language models retain or forget factual and relational knowledge when they are continuously edited over time.
The Art of Not Forgetting A Local Learning Architecture for Continual Learning
Ashmith Atmuri, Yashaswini Rao Bhogarajula
The paper presents CMP, a continual‑learning system that encodes inputs as sparse relational codes, stores them in a two‑tier competitive memory, and learns via local updates witho…
Regularizing modality contribution drift in multimodal continual learning
Zhen Zhang, Jielei Chu, Bin Liu +1
The paper identifies a decision-level shift called Modality Contribution Drift in multimodal continual learning and introduces a regularization method (CMCDR) that preserves modali…
GeoMFD: Continual Drone-View Geo-Localization with Geometry-Aware Adapter and Margin-Field Distillation
Zhongwei Chen, Hai-jun Rong, Tao Zhang +4
The paper introduces GeoMFD, a method that enables a single drone-view geo-localization model to continuously adapt to new environments while preserving its cross-view geometric re…
Knowledge-Aware Evolution for Task-Free Streaming Federated Continual Learning with Arbitrary Class Overlap
Sixing Tan, Xianmin Liu
The paper introduces FedKACE, a method for federated continual learning that works without task labels and handles streaming data with overlapping classes by adaptively switching i…
Gate-Zero Growth: A Geometric Framework for Function-Preserving Continual Learning
Dante Lok
The paper proposes gate-zero growth, a function‑preserving operator that adds new residual blocks via a zero‑initialized gate, enabling controlled function drift and near‑zero forg…
Mixtures of SubExperts for Large Language Continual Learning
Haeyong Kang, Hee Suk Yoon, Dahua Feng +1
The paper proposes Mixtures of SubExperts (MoSEs), a modular and sparse extension to transformer layers that uses lightweight sub-modules and a learned routing function to let larg…
NORACL: Neurogenesis for Oracle-free Resource-Adaptive Continual Learning
Karthik Charan Raghunathan, Christian Metzner, Laura Kriener +1
The paper introduces NORACL, a continual‑learning method that dynamically expands a neural network when representational or plasticity saturation is detected, eliminating the need…
Grow-Prune-Freeze Networks: Adaptive & Continual Learning Technique for Olfactory Navigation
Kordel K. France, Ovidiu Daescu
The paper proposes Grow‑Prune‑Freeze (GPF) networks, an adaptive continual‑learning framework that dynamically grows, prunes, and freezes early policy layers to enable robots to na…
AlphaWiSE: Adaptive Weight Interpolation for Continual Multimodal Representation Learning
Sarthak Jain, Qiran Hu, Zhen Zhu +1
The paper introduces AlphaWiSE, a post‑hoc weight‑space interpolation technique that combines two frozen checkpoints with learned scalar coefficients to improve continual learning…
Local Redundancy: An Information-Theoretic Measure of Plasticity from Synthetic Memorization
Jiaxuan Cheng
The paper introduces local redundancy, an information‑theoretic measure of neural network plasticity derived from universal compression theory, and shows that a computable lower bo…
Do Agent Optimizers Compound? A Continual-Learning Evaluation on Terminal-Bench 2.0
Wenxiao Wang, Priyatham Kattakinda, Soheil Feizi
The paper evaluates whether gains from agent-optimization methods compound over successive optimization phases in a continual‑learning setting, using hard tasks from Terminal‑Bench…
Can a Language Model Learn Facts Continually in Its Weights?
Charles O'Neill
The paper investigates whether language models can continuously acquire and retain factual knowledge directly in their weights, examining how different training styles affect reten…
Mistake gating leads to energy and memory efficient continual learning
Aaron Pache, Mark CW van Rossum
The paper introduces memorized mistake‑gated learning, a biologically inspired rule that updates neural network weights only when classification errors occur, cutting the number of…
Test-Time Learning with an Evolving Library
Weijia Xu, Alessandro Sordoni, Chandan Singh +4
The paper introduces EvoLib, a test-time learning framework that lets large language models build, reuse, and evolve a shared library of knowledge abstractions across tasks without…
Traceback Translators Against Forgetting in Continual Fake Speech Detection
Enrico Gottardis, Mattia Tamiazzo, Simone Milani
The paper proposes a method that uses a domain‑translator network to map new fake‑speech data back into the feature space of an existing detector, allowing continual learning witho…
Attribution-Guided Continual Learning for Large Language Models
Yazheng Liu, Yuxuan Wan, Rui Xu +3
The paper introduces an attribution-guided continual learning framework for large language models that uses Layer-wise Relevance Propagation to identify important parameters and li…
Separating Expert Retention from Autonomous Source Inference in Raw-ECG-Replay-Free Continual ECG Deployment
Yufan Lu, Xinhui Liu, Chenyang Xu +2
The paper proposes a replay‑free continual learning framework for multi‑source ECG classification that keeps a frozen pretrained backbone, adds a new linear expert for each incomin…
RL Forgets! Towards Continual Policy Optimization
Mao-Lin Luo, Zhe-Xu Wang, Zi-Hao Zhou +4
The paper investigates catastrophic forgetting in continual post‑training of vision‑language models with reinforcement learning, introduces the MRCL benchmark, and proposes a repla…
CA-DGCL: Dynamic Graph Continual Learning via Condensation and Attachment
Tingxu Yan Ye Yuan
The paper introduces CA-DGCL, a framework that compresses past graph snapshots into compact representations and uses tensor decomposition to create stable node features, which are…
One search, two signals: results blend meaning (embedding similarity, so papers that never use your words still surface) with keyword matches on titles, abstracts and summaries. Free, no sign-in needed.