PhD student, Stanford University
1 paper at NeurIPS 2025
We parallelize MCMC over the sequence, yielding more than order of magnitude wall-clock speedup on sampling.