{"paper":{"title":"Score-Repellent Monte Carlo: Toward Efficient Non-Markovian Sampler with Constant Memory in General State Spaces","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Score-Repellent Monte Carlo reduces asymptotic sampling variance as O(1/α) using constant-memory history summaries in general state spaces.","cross_cats":["stat.CO","stat.ML"],"primary_cat":"cs.LG","authors_text":"Bohyung Han, Do Young Eun, Geeho Kim, Jie Hu, Jinyoung Choi, Lingyun Chen","submitted_at":"2026-04-24T18:39:50Z","abstract_excerpt":"History-dependent sampling can reduce long-run Monte Carlo variance by discouraging redundant revisits, but existing schemes typically encode history through empirical measure on finite state spaces, which is infeasible in high-dimensional discrete configuration spaces or ill-posed in continuous domains. We propose Score-Repellent Monte Carlo (SRMC) framework that summarizes trajectory history by a running average of score evaluations in $\\mathbb{R}^d$, where $d$ is the dimension of the score and state representation. This history is converted into a surrogate target through an exponential sco"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"We further identify regimes in which the asymptotic covariance decreases as α increases, with scaling O(1/α), extending the near-zero-variance effect of finite-state history-dependent samplers to general state spaces with constant memory.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The assumptions required for the stochastic approximation analysis with controlled Markovian noise hold for the chosen base kernel and target distribution, allowing the coupled history recursion and estimators to converge as claimed.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"SRMC creates a history-dependent surrogate target via exponential tilt of a running score average, enabling non-Markovian sampling with O(1/alpha) asymptotic variance reduction and constant memory in general state spaces.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Score-Repellent Monte Carlo reduces asymptotic sampling variance as O(1/α) using constant-memory history summaries in general state spaces.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"7a0aeffd0b25af34034a695f5e8d32aaad374ccffecf05250b974843aa2b9243"},"source":{"id":"2604.22948","kind":"arxiv","version":2},"verdict":{"id":"a34124ac-69a3-475d-afbb-408558e183d5","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-08T12:06:26.036355Z","strongest_claim":"We further identify regimes in which the asymptotic covariance decreases as α increases, with scaling O(1/α), extending the near-zero-variance effect of finite-state history-dependent samplers to general state spaces with constant memory.","one_line_summary":"SRMC creates a history-dependent surrogate target via exponential tilt of a running score average, enabling non-Markovian sampling with O(1/alpha) asymptotic variance reduction and constant memory in general state spaces.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The assumptions required for the stochastic approximation analysis with controlled Markovian noise hold for the chosen base kernel and target distribution, allowing the coupled history recursion and estimators to converge as claimed.","pith_extraction_headline":"Score-Repellent Monte Carlo reduces asymptotic sampling variance as O(1/α) using constant-memory history summaries in general state spaces."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.22948/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"ai_meta_artifact","ran_at":"2026-05-21T09:44:06.387683Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T23:35:33.264356Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"b882050bd2c69eca268b00f46ba19a6e7e6f360b3c4a74f0ff44764abac12647"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}