10 papers
Amortising Bayesian Experimental Design for Sequential Information Gathering in LLMs
Jakob Hartmann, James Harvey, Jhonathan Navott +5
Large language models (LLMs) exhibit strong reasoning and world-knowledge capabilities, yet often struggle to gather information effectively across the multi-turn interactions requ…
Stabilizing Policy Gradients for Sample-Efficient Reinforcement Learning in LLM Reasoning
Luckeciano C. Melo, Alessandro Abate, Yarin Gal
Reinforcement Learning, particularly through policy gradient methods, has played a central role in enabling reasoning capabilities of Large Language Models. However, the optimizati…
Partial Reasoning in Language Models: Search and Refinement Guided by Uncertainty
Murilo da Luz, Bruno Brandão, Luana Martins +4
The use of Large Language Models (LLMs) for reasoning and planning tasks has drawn increasing attention in Artificial Intelligence research. Despite their remarkable progress, thes…
Iterative Deployment Improves Planning Skills in LLMs
Augusto B. Corrêa, Yoav Gelberg, Luckeciano C. Melo +3
We show that iterative deployment of large language models (LLMs), each fine-tuned on data carefully curated by users from the previous models' deployment, can significantly change…
Learning Without Critics? Revisiting GRPO in Classical Reinforcement Learning Environments
Bryan L. M. de Oliveira, Felipe V. Frujeri, Marcos P. C. M. Queiroz +3
Group Relative Policy Optimization (GRPO) has emerged as a scalable alternative to Proximal Policy Optimization (PPO) by eliminating the learned critic and instead estimating advan…
Temporal-Difference Variational Continual Learning
Luckeciano C. Melo, Alessandro Abate, Yarin Gal
Machine Learning models in real-world applications must continuously learn new tasks to adapt to shifts in the data-generating distribution. Yet, for Continual Learning (CL), model…