Learning to Plan with Reinforcement Learning
Abstract
This thesis investigates the domain of learning to plan within reinforcement learning. Traditional planning algorithms, such as Monte Carlo Tree Search, are typically handcrafted and lack the flexibility inherent in biological planning systems. To address this limitation, the Thinker algorithm is proposed. A central theme of this work is the shift in focus from designing specialized planning algorithms to augmenting the environment. By treating planning as an explicit action within an augmented environment, the agent is enabled to learn to plan using standard model-free reinforcement learning. Additionally, this thesis examines the interpretability of planning behaviour of RL agents trained by different algorithms. Finally, this thesis extends the Thinker algorithm to Large Language Models trained on mathematical reasoning tasks, demonstrating that similarly structured augmented environments significantly reduce computational costs while maintaining competitive performance on reasoning tasks.
Keywords
Not available
Specialisations
Similar theses
- Learning to adapt: From structured domain adaptation to efficient in-context learning for large language models Zihao Xu · Rutgers University · 2026
- Causal Structure Learning for Robust and Explainable Artificial Intelligence Brandon David Mossop · Queen's University Library · 2026
- AI-Generated Image Detection Using Multimodal Deep Learning Mamadou Keïta · 2026
- Multimodal and explainable Learning using Fusion Architecture Through Exploration (FATE) approach Otmane Amel · 2026
- On large deviations in probabilistic deep learning and generative modeling Nilsson, Viktor · KTH Royal Institute of Technology · 2025
- Explainable AI: Enhancing Machine Learning Model Interpretability with Generative AI Castelhano, Inês Dinis · 2025
