Imagen de Portada

Heads collapse, features stay: Why Replay needs big buffers

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Lanzillotta, Giulia, Meier, Damiano, Hofmann, Thomas
Formato:	Preprint
Publicado:	2025
Materias:	Machine Learning Artificial Intelligence
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

The Importance of Being Lazy: Scaling Limits of Continual Learning
por: Graldi, Jacopo, et al.
Publicado: (2025)

Revisiting Knowledge Distillation: The Hidden Role of Dataset Size
por: Lanzillotta, Giulia, et al.
Publicado: (2025)

Local vs Global continual learning
por: Lanzillotta, Giulia, et al.
Publicado: (2024)

Reactivation: Empirical NTK Dynamics Under Task Shifts
por: Liu, Yuzhi, et al.
Publicado: (2025)

Replay-buffer engineering for noise-robust quantum circuit optimization
por: Kundu, Akash, et al.
Publicado: (2026)

Barriers for Learning in an Evolving World: Mathematical Understanding of Loss of Plasticity
por: Joudaki, Amir, et al.
Publicado: (2025)

Enhancing interpretability of rule-based classifiers through feature graphs
por: Sirocchi, Christel, et al.
Publicado: (2025)

Prioritized Trajectory Replay: A Replay Memory for Data-driven Reinforcement Learning
por: Liu, Jinyi, et al.
Publicado: (2023)

Automated data processing and feature engineering for deep learning and big data applications: a survey
por: Mumuni, Alhassan, et al.
Publicado: (2024)

Experience Replay with Random Reshuffling
por: Fujita, Yasuhiro
Publicado: (2025)

VLM-Guided Experience Replay
por: Sharony, Elad, et al.
Publicado: (2026)

Investigating the Interplay of Prioritized Replay and Generalization
por: Panahi, Parham Mohammad, et al.
Publicado: (2024)

ARROW: Augmented Replay for RObust World models
por: Alyahya, Abdulaziz, et al.
Publicado: (2026)

Better Generative Replay for Continual Federated Learning
por: Qi, Daiqing, et al.
Publicado: (2023)

Adaptive Replay Buffer for Offline-to-Online Reinforcement Learning
por: Song, Chihyeon, et al.
Publicado: (2025)

Mastering the Game of Go with Self-play Experience Replay
por: Liu, Jingbin, et al.
Publicado: (2026)

ROER: Regularized Optimal Experience Replay
por: Li, Changling, et al.
Publicado: (2024)

When Does Non-Uniform Replay Matter in Reinforcement Learning?
por: Korniak, Michal, et al.
Publicado: (2026)

Finite-Time Analysis of Temporal Difference Learning with Experience Replay
por: Lim, Han-Dong, et al.
Publicado: (2023)

Efficient Diversity-based Experience Replay for Deep Reinforcement Learning
por: Zhao, Kaiyan, et al.
Publicado: (2024)

Generalized Back-Stepping Experience Replay in Sparse-Reward Environments
por: Lyu, Guwen, et al.
Publicado: (2024)

An Efficient Replay for Class-Incremental Learning with Pre-trained Models
por: Yin, Weimin, et al.
Publicado: (2024)

Data-Free Generative Replay for Class-Incremental Learning on Imbalanced Data
por: Younis, Sohaib, et al.
Publicado: (2024)

R^3: Replay, Reflection, and Ranking Rewards for LLM Reinforcement Learning
por: Jiang, Zhizheng, et al.
Publicado: (2026)

CORE: Mitigating Catastrophic Forgetting in Continual Learning through Cognitive Replay
por: Zhang, Jianshu, et al.
Publicado: (2024)

Prioritized Replay for RL Post-training
por: Fatemi, Mehdi
Publicado: (2026)

Teaching AI to Remember: Insights from Brain-Inspired Replay in Continual Learning
por: Kim, Jina
Publicado: (2025)

EFRame: Deeper Reasoning via Exploration-Filter-Replay Reinforcement Learning Framework
por: Wang, Chen, et al.
Publicado: (2025)

Continual Offline Reinforcement Learning via Diffusion-based Dual Generative Replay
por: Liu, Jinmei, et al.
Publicado: (2024)

Robust and Resource-Efficient Data-Free Knowledge Distillation by Generative Pseudo Replay
por: Binici, Kuluhan, et al.
Publicado: (2022)

Hindsight Preference Replay Improves Preference-Conditioned Multi-Objective Reinforcement Learning
por: Shianifar, Jonaid, et al.
Publicado: (2026)

Provable Effects of Data Replay in Continual Learning: A Feature Learning Perspective
por: Ding, Meng, et al.
Publicado: (2026)

Rooted Absorbed Prefix Trajectory Balance with Submodular Replay for GFlowNet Training
por: Wang, Xi, et al.
Publicado: (2026)

PAGE: Domain-Incremental Adaptation with Past-Agnostic Generative Replay for Smart Healthcare
por: Li, Chia-Hao, et al.
Publicado: (2024)

ReplaySCM: A Benchmark for Executable Causal Mechanism Induction from Interventions
por: Batzoglou, Serafim
Publicado: (2026)

Defense without Forgetting: Continual Adversarial Defense with Anisotropic & Isotropic Pseudo Replay
por: Zhou, Yuhang, et al.
Publicado: (2024)

Dual-LoRA and Quality-Enhanced Pseudo Replay for Multimodal Continual Food Learning
por: Wu, Xinlan, et al.
Publicado: (2025)

The Effectiveness of Approximate Regularized Replay for Efficient Supervised Fine-Tuning of Large Language Models
por: Riemer, Matthew, et al.
Publicado: (2025)

On the Convergence of Experience Replay in Policy Optimization: Characterizing Bias, Variance, and Finite-Time Convergence
por: Zheng, Hua, et al.
Publicado: (2021)

Revisiting Softmax Masking: Stop Gradient for Enhancing Stability in Replay-based Continual Learning
por: Kim, Hoyong, et al.
Publicado: (2023)