{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:YL4X2YJMG6YFVKUQQOG4W2MMYF","short_pith_number":"pith:YL4X2YJM","schema_version":"1.0","canonical_sha256":"c2f97d612c37b05aaa90838dcb698cc15346d0e0be6a50bccad7ae8f1e1951dd","source":{"kind":"arxiv","id":"2501.09685","version":2},"attestation_state":"computed","paper":{"title":"Inference-Time Alignment in Diffusion Models with Reward-Guided Generation: Tutorial and Review","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG","q-bio.QM","stat.ML"],"primary_cat":"cs.AI","authors_text":"Aviv Regev, Chenyu Wang, Masatoshi Uehara, Sergey Levine, Tommaso Biancalani, Xiner Li, Yulai Zhao","submitted_at":"2025-01-16T17:37:35Z","abstract_excerpt":"This tutorial provides an in-depth guide on inference-time guidance and alignment methods for optimizing downstream reward functions in diffusion models. While diffusion models are renowned for their generative modeling capabilities, practical applications in fields such as biology often require sample generation that maximizes specific metrics (e.g., stability, affinity in proteins, closeness to target structures). In these scenarios, diffusion models can be adapted not only to generate realistic samples but also to explicitly maximize desired measures at inference time without fine-tuning. T"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2501.09685","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-01-16T17:37:35Z","cross_cats_sorted":["cs.LG","q-bio.QM","stat.ML"],"title_canon_sha256":"d723f75af1ab3cd56aea760ee7ce313f7df7bfc99ab56def971980dcda36bf2b","abstract_canon_sha256":"fd16a8c5ec247156c6bd48640671f1df1f9b875dd147599446bdf1b160d32931"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:02:58.860947Z","signature_b64":"Iz8rFpF96tFVIh21w74HfmtaRoknzJIkS+OWKPvOkn9gukSGZtxo3QAz78ml6fhQdBDLVZo0ejmrl6txJv7yDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c2f97d612c37b05aaa90838dcb698cc15346d0e0be6a50bccad7ae8f1e1951dd","last_reissued_at":"2026-07-05T10:02:58.860459Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:02:58.860459Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Inference-Time Alignment in Diffusion Models with Reward-Guided Generation: Tutorial and Review","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG","q-bio.QM","stat.ML"],"primary_cat":"cs.AI","authors_text":"Aviv Regev, Chenyu Wang, Masatoshi Uehara, Sergey Levine, Tommaso Biancalani, Xiner Li, Yulai Zhao","submitted_at":"2025-01-16T17:37:35Z","abstract_excerpt":"This tutorial provides an in-depth guide on inference-time guidance and alignment methods for optimizing downstream reward functions in diffusion models. While diffusion models are renowned for their generative modeling capabilities, practical applications in fields such as biology often require sample generation that maximizes specific metrics (e.g., stability, affinity in proteins, closeness to target structures). In these scenarios, diffusion models can be adapted not only to generate realistic samples but also to explicitly maximize desired measures at inference time without fine-tuning. T"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.09685","kind":"arxiv","version":2},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2501.09685/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2501.09685","created_at":"2026-07-05T10:02:58.860511+00:00"},{"alias_kind":"arxiv_version","alias_value":"2501.09685v2","created_at":"2026-07-05T10:02:58.860511+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.09685","created_at":"2026-07-05T10:02:58.860511+00:00"},{"alias_kind":"pith_short_12","alias_value":"YL4X2YJMG6YF","created_at":"2026-07-05T10:02:58.860511+00:00"},{"alias_kind":"pith_short_16","alias_value":"YL4X2YJMG6YFVKUQ","created_at":"2026-07-05T10:02:58.860511+00:00"},{"alias_kind":"pith_short_8","alias_value":"YL4X2YJM","created_at":"2026-07-05T10:02:58.860511+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":23,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2604.17415","citing_title":"Reward Score Matching: Unifying Reward-based Fine-tuning for Flow and Diffusion Models","ref_index":51,"is_internal_anchor":true},{"citing_arxiv_id":"2606.08802","citing_title":"Active Flow Expansion for Out-of-Distribution Discovery: from Theory to Molecules","ref_index":54,"is_internal_anchor":false},{"citing_arxiv_id":"2606.08375","citing_title":"Few-step Cofolding with All-Atom Flow Maps","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2606.02884","citing_title":"Are we really tilting? The mechanics of reward guidance in flow and diffusion models","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2606.00267","citing_title":"StressDream: Steering Video World Models for Robust Policy Evaluation and Improvement","ref_index":83,"is_internal_anchor":false},{"citing_arxiv_id":"2604.27147","citing_title":"How to Guide Your Flow: Few-Step Alignment via Flow Map Reward Guidance","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2605.18745","citing_title":"SURGE: Approximation and Training Free Particle Filter for Diffusion Surrogate","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2605.25123","citing_title":"Inference-Time Alignment of Diffusion Models via Trust-Region Iterative Twisted Sequential Monte Carlo","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2605.26552","citing_title":"Aligning Few-Step Generative Models by Amortizing Sample-based Variational Inference","ref_index":95,"is_internal_anchor":false},{"citing_arxiv_id":"2605.30991","citing_title":"Parallel Tempering Initial Sampling in Inference-Time Reward Alignment","ref_index":42,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07220","citing_title":"On the Robustness of Distribution Support under Diffusion Guidance","ref_index":86,"is_internal_anchor":false},{"citing_arxiv_id":"2505.19075","citing_title":"Universal Reasoner: A Single, Composable Plug-and-Play Reasoner for Frozen LLMs","ref_index":38,"is_internal_anchor":false},{"citing_arxiv_id":"2604.27147","citing_title":"How to Guide Your Flow: Few-Step Alignment via Flow Map Reward Guidance","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2605.18745","citing_title":"SURGE: Approximation and Training Free Particle Filter for Diffusion Surrogate","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17850","citing_title":"Simple Approximation and Derivative Free Inference-Time Scaling for Diffusion Models via Sequential Monte Carlo on Path Measures","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2510.06637","citing_title":"Control-Augmented Autoregressive Diffusion for Data Assimilation","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2510.12916","citing_title":"Efficient Inference for Coupled Hidden Markov Models in Continuous Time and Discrete Space","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2510.20206","citing_title":"RAPO++: Cross-Stage Prompt Optimization for Text-to-Video Generation via Data Alignment and Test-Time Scaling","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2512.23532","citing_title":"Iterative Inference-time Scaling with Adaptive Frequency Steering for Image Super-Resolution","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2604.27147","citing_title":"How to Guide Your Flow: Few-Step Alignment via Flow Map Reward Guidance","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07220","citing_title":"On the Robustness of Distribution Support under Diffusion Guidance","ref_index":86,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07456","citing_title":"Inference-Time Attribute Distribution Alignment for Unconditional Diffusion","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07253","citing_title":"LENS: Low-Frequency Eigen Noise Shaping for Efficient Diffusion Sampling","ref_index":37,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YL4X2YJMG6YFVKUQQOG4W2MMYF","json":"https://pith.science/pith/YL4X2YJMG6YFVKUQQOG4W2MMYF.json","graph_json":"https://pith.science/api/pith-number/YL4X2YJMG6YFVKUQQOG4W2MMYF/graph.json","events_json":"https://pith.science/api/pith-number/YL4X2YJMG6YFVKUQQOG4W2MMYF/events.json","paper":"https://pith.science/paper/YL4X2YJM"},"agent_actions":{"view_html":"https://pith.science/pith/YL4X2YJMG6YFVKUQQOG4W2MMYF","download_json":"https://pith.science/pith/YL4X2YJMG6YFVKUQQOG4W2MMYF.json","view_paper":"https://pith.science/paper/YL4X2YJM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2501.09685&json=true","fetch_graph":"https://pith.science/api/pith-number/YL4X2YJMG6YFVKUQQOG4W2MMYF/graph.json","fetch_events":"https://pith.science/api/pith-number/YL4X2YJMG6YFVKUQQOG4W2MMYF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YL4X2YJMG6YFVKUQQOG4W2MMYF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YL4X2YJMG6YFVKUQQOG4W2MMYF/action/storage_attestation","attest_author":"https://pith.science/pith/YL4X2YJMG6YFVKUQQOG4W2MMYF/action/author_attestation","sign_citation":"https://pith.science/pith/YL4X2YJMG6YFVKUQQOG4W2MMYF/action/citation_signature","submit_replication":"https://pith.science/pith/YL4X2YJMG6YFVKUQQOG4W2MMYF/action/replication_record"}},"created_at":"2026-07-05T10:02:58.860511+00:00","updated_at":"2026-07-05T10:02:58.860511+00:00"}