Aaron Courville

4 papers Mila, U. Montreal

• Compositional Discrete Latent Code for High Fidelity, Productive Diffusion Models
• Measure gradients, not activations! Enhancing neuronal activity in deep reinforcement learning
• Mixture-of-Recursions: Learning Dynamic Recursive Depths for Adaptive Token-Level Computation
• Stable Gradients for Stable Learning at Scale in Deep Reinforcement Learning