{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:HG6YFJNKAEZ76FAIZJ67CZ2CHC","short_pith_number":"pith:HG6YFJNK","schema_version":"1.0","canonical_sha256":"39bd82a5aa0133ff1408ca7df1674238868c782a0dbcc6a13b456ad03d73c4e0","source":{"kind":"arxiv","id":"2603.17919","version":2},"attestation_state":"computed","paper":{"title":"Training Diffusion Language Models for Black-Box Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CE","authors_text":"Can Chen, Christopher Pal, Haolun Wu, Jiayao Gu, Xue Liu, Ye Yuan, Zipeng Sun","submitted_at":"2026-03-18T16:56:23Z","abstract_excerpt":"We study offline black-box optimization (BBO), aiming to discover improved designs from an offline dataset of designs and labels, a problem common in robotics, DNA, and materials science with limited labeled samples. While recent work applies autoregressive LLMs to BBO by formatting tasks as natural-language prompts, their left-to-right design generation struggles to capture the strong bidirectional dependencies inherent in design problems. To address this, we propose adapting diffusion LLMs to offline BBO to leverage their bidirectional modeling capabilities. However, a domain gap exists betw"},"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":"2603.17919","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CE","submitted_at":"2026-03-18T16:56:23Z","cross_cats_sorted":[],"title_canon_sha256":"299e2a5f9131429396092f05ec40857ef80bd36ae35540da994b17ae08bdc762","abstract_canon_sha256":"d60d90ffa8161e2e385a5a4e0a993546ac480caaf3ba613c55f312b7c825af9a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-26T01:03:28.264906Z","signature_b64":"jlTAUbUG881yCLfeMK0fVoDZTtRoyrnujVYugp/D/F1KEoNxGVWlP3zxN7LBknuKiirWgOhkBDdd/O5/3G4bCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"39bd82a5aa0133ff1408ca7df1674238868c782a0dbcc6a13b456ad03d73c4e0","last_reissued_at":"2026-05-26T01:03:28.263789Z","signature_status":"signed_v1","first_computed_at":"2026-05-26T01:03:28.263789Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Training Diffusion Language Models for Black-Box Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CE","authors_text":"Can Chen, Christopher Pal, Haolun Wu, Jiayao Gu, Xue Liu, Ye Yuan, Zipeng Sun","submitted_at":"2026-03-18T16:56:23Z","abstract_excerpt":"We study offline black-box optimization (BBO), aiming to discover improved designs from an offline dataset of designs and labels, a problem common in robotics, DNA, and materials science with limited labeled samples. While recent work applies autoregressive LLMs to BBO by formatting tasks as natural-language prompts, their left-to-right design generation struggles to capture the strong bidirectional dependencies inherent in design problems. To address this, we propose adapting diffusion LLMs to offline BBO to leverage their bidirectional modeling capabilities. However, a domain gap exists betw"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2603.17919","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/2603.17919/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":"2603.17919","created_at":"2026-05-26T01:03:28.263916+00:00"},{"alias_kind":"arxiv_version","alias_value":"2603.17919v2","created_at":"2026-05-26T01:03:28.263916+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.17919","created_at":"2026-05-26T01:03:28.263916+00:00"},{"alias_kind":"pith_short_12","alias_value":"HG6YFJNKAEZ7","created_at":"2026-05-26T01:03:28.263916+00:00"},{"alias_kind":"pith_short_16","alias_value":"HG6YFJNKAEZ76FAI","created_at":"2026-05-26T01:03:28.263916+00:00"},{"alias_kind":"pith_short_8","alias_value":"HG6YFJNK","created_at":"2026-05-26T01:03:28.263916+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2605.22144","citing_title":"One Sentence, One Drama: Personalized Short-Form Drama Generation via Multi-Agent Systems","ref_index":37,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HG6YFJNKAEZ76FAIZJ67CZ2CHC","json":"https://pith.science/pith/HG6YFJNKAEZ76FAIZJ67CZ2CHC.json","graph_json":"https://pith.science/api/pith-number/HG6YFJNKAEZ76FAIZJ67CZ2CHC/graph.json","events_json":"https://pith.science/api/pith-number/HG6YFJNKAEZ76FAIZJ67CZ2CHC/events.json","paper":"https://pith.science/paper/HG6YFJNK"},"agent_actions":{"view_html":"https://pith.science/pith/HG6YFJNKAEZ76FAIZJ67CZ2CHC","download_json":"https://pith.science/pith/HG6YFJNKAEZ76FAIZJ67CZ2CHC.json","view_paper":"https://pith.science/paper/HG6YFJNK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2603.17919&json=true","fetch_graph":"https://pith.science/api/pith-number/HG6YFJNKAEZ76FAIZJ67CZ2CHC/graph.json","fetch_events":"https://pith.science/api/pith-number/HG6YFJNKAEZ76FAIZJ67CZ2CHC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HG6YFJNKAEZ76FAIZJ67CZ2CHC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HG6YFJNKAEZ76FAIZJ67CZ2CHC/action/storage_attestation","attest_author":"https://pith.science/pith/HG6YFJNKAEZ76FAIZJ67CZ2CHC/action/author_attestation","sign_citation":"https://pith.science/pith/HG6YFJNKAEZ76FAIZJ67CZ2CHC/action/citation_signature","submit_replication":"https://pith.science/pith/HG6YFJNKAEZ76FAIZJ67CZ2CHC/action/replication_record"}},"created_at":"2026-05-26T01:03:28.263916+00:00","updated_at":"2026-05-26T01:03:28.263916+00:00"}