Continual Learning in Sequential Tasks

Project Overview

Investigating approaches to enable machine learning systems to learn new tasks sequentially while maintaining performance on previously learned tasks.

Research Focus

This work explores methods for continual learning in reinforcement learning settings, addressing the challenge of catastrophic forgetting when learning multiple tasks sequentially.

Key Areas:

• Catastrophic forgetting mitigation
• Multi-task learning
• Sequential task learning in RL

Target Publication

ICLR 2026 - Research in progress on continual learning methods for sequential tasks.