Two papers from LiteRL have been accepted at the Reinforcement Learning Conference (RLC 2025) and published in the Reinforcement Learning Journal (RLJ).
Learning to Explore in Diverse Reward Settings via Temporal-Difference-Error Maximization
S. Griesbach, C. D’Eramo
SEE (Stable Error-seeking Exploration) explores by seeking out surprising transitions. It works across dense, sparse and even exploration-adverse rewards, without retuning hyperparameters. Read the paper →
Eau De Q-Network: Adaptive Distillation of Neural Networks in Deep Reinforcement Learning
T. Vincent, T. Faust, Y. Tripathi, J. Peters, C. D’Eramo
EauDeQN makes agents sparse at their own learning pace, discovering how far a network can be pruned instead of fixing the final sparsity in advance. Read the paper →
Congratulations to all authors! See all our work on the Publications page.