Continual Policy Consolidation for Lifelong Robot Learning
cs.LG
Submitted: 2026-01-30
Updated: 2026-09-06
Comments: 22 pages
Code: https://github.com/rainx0r/metaworld-algorithms
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- Expert Gate: Lifelong Learning with a Network of Experts
- Theory on Mixture-of-Experts in Continual Learning
- Prioritized Trajectory Replay: A Replay Memory for Data-driven Reinforcement Learning
- Net2Net: Accelerating Learning via Knowledge Transfer
- Decoupled Weight Decay Regularization
- Lifelong Language Pretraining with Distribution-Specialized Experts
- Maintaining Plasticity in Deep Continual Learning
- One Model for All Tasks: Leveraging Efficient World Models in Multi-Task Planning
- Multi-Task Reinforcement Learning with Mixture of Orthogonal Experts
- Comparing Kullback-Leibler Divergence and Mean Squared Error Loss in Knowledge Distillation
- A Generalist Agent
- InfoNCE: Identifying the Gap Between Theory and Practice
- Proximal Policy Optimization Algorithms
- Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer
- Continual Learning and Catastrophic Forgetting
- 1000 Layer Networks for Self-Supervised RL: Scaling Depth Can Enable New Goal-Reaching Capabilities
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: Local-Consistency Optimisation for Bayesian Neural Networks
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks