{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:RN55OZIG5BJCD2MJ2GJ343S7NV","short_pith_number":"pith:RN55OZIG","schema_version":"1.0","canonical_sha256":"8b7bd76506e85221e989d193be6e5f6d60fbc6390735332fe8237d397a968535","source":{"kind":"arxiv","id":"2506.20701","version":1},"attestation_state":"computed","paper":{"title":"Diffusion Tree Sampling: Scalable inference-time alignment of diffusion models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Kusha Sareen, Mohammad Pedramfar, Siamak Ravanbakhsh, Vineet Jain","submitted_at":"2025-06-25T17:59:10Z","abstract_excerpt":"Adapting a pretrained diffusion model to new objectives at inference time remains an open problem in generative modeling. Existing steering methods suffer from inaccurate value estimation, especially at high noise levels, which biases guidance. Moreover, information from past runs is not reused to improve sample quality, resulting in inefficient use of compute. Inspired by the success of Monte Carlo Tree Search, we address these limitations by casting inference-time alignment as a search problem that reuses past computations. We introduce a tree-based approach that samples from the reward-alig"},"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":"2506.20701","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-06-25T17:59:10Z","cross_cats_sorted":["cs.AI","stat.ML"],"title_canon_sha256":"98ca537090e3c85965624fe9d50de8bbb7ab4e577f28806fac9b601b5c2b9be2","abstract_canon_sha256":"6f59b076b283e557ee6803da8a6017025e9906e746d2784b599f26e26e4f9603"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:27:32.659987Z","signature_b64":"Ht3jVAb+BLACxGNYgoExzGUp5gnEIFAlmjVQDrSlA8UtxyOIA5VumFzgcuigTUzNUnVvY1UpEMltUDI5drS0Bw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8b7bd76506e85221e989d193be6e5f6d60fbc6390735332fe8237d397a968535","last_reissued_at":"2026-07-05T11:27:32.659492Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:27:32.659492Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Diffusion Tree Sampling: Scalable inference-time alignment of diffusion models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Kusha Sareen, Mohammad Pedramfar, Siamak Ravanbakhsh, Vineet Jain","submitted_at":"2025-06-25T17:59:10Z","abstract_excerpt":"Adapting a pretrained diffusion model to new objectives at inference time remains an open problem in generative modeling. Existing steering methods suffer from inaccurate value estimation, especially at high noise levels, which biases guidance. Moreover, information from past runs is not reused to improve sample quality, resulting in inefficient use of compute. Inspired by the success of Monte Carlo Tree Search, we address these limitations by casting inference-time alignment as a search problem that reuses past computations. We introduce a tree-based approach that samples from the reward-alig"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.20701","kind":"arxiv","version":1},"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/2506.20701/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":"2506.20701","created_at":"2026-07-05T11:27:32.659554+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.20701v1","created_at":"2026-07-05T11:27:32.659554+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.20701","created_at":"2026-07-05T11:27:32.659554+00:00"},{"alias_kind":"pith_short_12","alias_value":"RN55OZIG5BJC","created_at":"2026-07-05T11:27:32.659554+00:00"},{"alias_kind":"pith_short_16","alias_value":"RN55OZIG5BJCD2MJ","created_at":"2026-07-05T11:27:32.659554+00:00"},{"alias_kind":"pith_short_8","alias_value":"RN55OZIG","created_at":"2026-07-05T11:27:32.659554+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":7,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.08375","citing_title":"Few-step Cofolding with All-Atom Flow Maps","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2607.01144","citing_title":"Sequentially-Controlled Interactive Multi-Particle Flow-Maps for Online Feedback-Driven Search","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2511.08425","citing_title":"HardFlow: Hard-Constrained Sampling for Flow-Matching Models via Trajectory Optimization","ref_index":48,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06779","citing_title":"VASR: Variance-Aware Systematic Resampling for Reward-Guided Diffusion","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09433","citing_title":"Offline Preference Optimization for Rectified Flow with Noise-Tracked Pairs","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2604.22212","citing_title":"Multimodal Diffusion to Mutually Enhance Polarized Light and Low Resolution EBSD Data","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06779","citing_title":"VASR: Variance-Aware Systematic Resampling for Reward-Guided Diffusion","ref_index":22,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/RN55OZIG5BJCD2MJ2GJ343S7NV","json":"https://pith.science/pith/RN55OZIG5BJCD2MJ2GJ343S7NV.json","graph_json":"https://pith.science/api/pith-number/RN55OZIG5BJCD2MJ2GJ343S7NV/graph.json","events_json":"https://pith.science/api/pith-number/RN55OZIG5BJCD2MJ2GJ343S7NV/events.json","paper":"https://pith.science/paper/RN55OZIG"},"agent_actions":{"view_html":"https://pith.science/pith/RN55OZIG5BJCD2MJ2GJ343S7NV","download_json":"https://pith.science/pith/RN55OZIG5BJCD2MJ2GJ343S7NV.json","view_paper":"https://pith.science/paper/RN55OZIG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.20701&json=true","fetch_graph":"https://pith.science/api/pith-number/RN55OZIG5BJCD2MJ2GJ343S7NV/graph.json","fetch_events":"https://pith.science/api/pith-number/RN55OZIG5BJCD2MJ2GJ343S7NV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/RN55OZIG5BJCD2MJ2GJ343S7NV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/RN55OZIG5BJCD2MJ2GJ343S7NV/action/storage_attestation","attest_author":"https://pith.science/pith/RN55OZIG5BJCD2MJ2GJ343S7NV/action/author_attestation","sign_citation":"https://pith.science/pith/RN55OZIG5BJCD2MJ2GJ343S7NV/action/citation_signature","submit_replication":"https://pith.science/pith/RN55OZIG5BJCD2MJ2GJ343S7NV/action/replication_record"}},"created_at":"2026-07-05T11:27:32.659554+00:00","updated_at":"2026-07-05T11:27:32.659554+00:00"}