SampleWiki
New results, open questions and their proofs. Read the mathematics; open the checked Lean when you need it.
Browse sampling regimes
Start with the target assumptions, oracle and accuracy notion. Compare results and open questions in that setting.
Research resources: guides, formal libraries and papers
A textbook-linked worked example
Ideal proximal chain
Selected research projects
Read a paper or case study, its proof route and the pieces verified so far.
Proximal BPS
High accuracy from a warm start: process, invariant kernels and convergence.
arXiv:2609.06905v1Statement and proof →Smoothed Picard HMC
Smoothing, Picard dynamics and the proxy warm start.
arXiv:2609.06906v1Statement and proof →Gaussian Cloud · Part I
Recursive simulation, conditional errors and Section 6 cost bookkeeping.
arXiv:2609.38710v1Statement and proof →Midpoint Langevin
Dimension-dependent upper bounds and separate oracle lower bounds.
arXiv:2610.06308v1Statement and proof →Gaussian Sampling Complexity
Exact-oracle lower bounds, Gaussian expectations and the laws behind the proof.
arXiv:2610.20492v1Statement and proof →Verified building blocks are not completed paper proofs.
How the warm start and high-accuracy sampler fit together →
Open mathematical questions
These are open in the literature, not merely missing Lean proofs.
General sampling lower bound
Matching lower bound not currently known in the pinned literature trail
Matching result remains open.
Matching PI/LSI oracle lower bound
Matching lower bound not currently known in the pinned literature trail
Matching result remains open.
Matching Hölder-model lower bound
Matching lower bound not currently known in the pinned literature trail
Matching result remains open.
Matching first-order lower bound
Matching lower bound not currently known in the pinned literature trail