Markov Chain Monte Carlo (MCMC) methods are fundamental for sampling from complex probability distributions, but many widely used algorithms either rely on gradients (like NUTS) and/or struggle with high-dimensional, multi-scale … Read More
Traditional methods like Gibbs sampling or randomized Kaczmarz rely heavily on specific coordinate systems, which can limit their efficiency—especially in ill-conditioned settings. But what happens when we step away from … Read More
Curious about the No-U-Turn Sampler and its performance in high-dimensional spaces? Watch my recent talk, where I present new insights into its reversibility and mixing time. There are still many … Read More
I’m thrilled to announce that I’ll be presenting at the Online Seminar on Monte Carlo on October 22 from 11:30 AM – 12:30 PM EST! I’ll be discussing recent work … Read More
The No-U-Turn Sampler (NUTS) is the go-to method for sampling in probabilistic programming languages like Stan, PyMC3, NIMBLE, Turing, and NumPyro. However, due to its recursive architecture, even proving its … Read More
Excited to share that I’ll be speaking at the Penn/Temple probability seminar on October 1st. Looking forward to discussing the recent progress we’ve made in understanding the reversibility of the … Read More
Excited to attend the upcoming Recent Advances and Future Directions for Sampling conference at Yale. With so much progress happening in MCMC and related areas, this event feels perfectly timed … Read More
Imagine a No-U-Turn Sampler that can adapt both its path-length and step-size on the fly, responding to the local geometry of the target distribution, while still preserving detailed balance. In … Read More
In a highly cited paper, Diaconis et al, (2000) lifted the symmetric random walk on the discrete circle by adding an auxiliary velocity variable that is flipped in each step … Read More
Anyone who has ever used Hamiltonian Monte Carlo samplers before has probably encountered the so-called tuning problem: in short, it is not at all obvious how to pick the algorithm’s … Read More