AI Paper Atlas

Research map / Probabilistic and scientific modeling / Variational and generative sampling

Geometry-Aware Score-Repellent Monte Carlo

Jie Hu, Lingyun Chen, Do-Young Eun

NeurIPS 2026 · Probabilistic and scientific modeling › Variational and generative sampling · approach: Hamiltonian Monte Carlo · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes a geometry-aware extension of Score-Repellent Monte Carlo (SRMC) that replaces the single scalar repellence strength with a direction-dependent one to handle steep and flat directions of the target.

Problem

Score-Repellent Monte Carlo tilts the target density with a single scalar repellence strength, which cannot simultaneously handle steep and flat directions of the target distribution in MCMC sampling.

Approach

Extends SRMC's score-based repellence factor exp(-alpha theta_n^T s(x)) to be geometry-aware so that repellence strength can adapt across steep and flat directions instead of using one scalar alpha.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
replace Euclidean alignment (SRMC) with Fisher-weighted alignment (GA-SRMC) on MALAMALA + SRMC · Gauss-2D kappa=100worst-coordinate ESS: 104 -> 452Improved
replace SRMC with GA-SRMC on MALAMALA + SRMC · MNIST BLRworst-coordinate ESS: 74 -> 501Improved
replace SRMC with GA-SRMC on MALAMALA + SRMC · Australian, Heart, Pima BLRworst-coordinate ESS: 16-32% higher ESSImproved
replace SRMC with GA-SRMC on FisherMALA (proposal already preconditioned)FisherMALA + SRMC · Gauss-2Dworst-coordinate ESS: 12459 -> 18425 (ratio 1.48)Improved
replace SRMC with GA-SRMC on FisherMALAFisherMALA + SRMC · Caravan BLRworst-coordinate ESS: 6381 -> 6022Hurt
replace SRMC with GA-SRMC on MALAMALA + SRMC · German BLRworst-coordinate ESS: 4663 -> 4173Hurt
replace SRMC with GA-SRMCMALA/FisherMALA + SRMC · all 8 targetsposterior-mean MSE: lower in all 16 full-Fisher comparisonsImproved
replace SRMC with GA-SRMC on near-isotropic targetMALA + SRMC · Ripley (kappa=1.3)posterior-mean MSE: 0.205 -> 0.198No clear effect
add GA-SRMC to MALA versus swapping MALA for FisherMALAMALA · all 8 targetsposterior-mean MSE: 2.45x on MNIST to 32.9x on Ripley lower than FisherMALAImproved
use smaller Fisher regularizer delta (0.01 instead of 10) (ablation)FisherMALA + GA-SRMC · 7 targets, delta sweepworst-coordinate ESS: Gauss-2D 10 -> 12695; Ripley 12027 -> 13958Improved
use diagonal instead of full Fisher estimate in GA-SRMC under MALA (ablation)MALA + GA-SRMC · coupled BLR posteriorsworst-coordinate ESSHurt
use diagonal Fisher estimate in history feedback only (MALA + diagonal GA-SRMC) (ablation)MALA + SRMC · all 8 targetsposterior-mean MSEImproved

Experimental setup

Baselines: Score-Repellent Monte Carlo

Builds on: Score-Repellent Monte Carlo

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy