{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:4YYJD5IFUYGAAPLCJ7D7BQAEYF","short_pith_number":"pith:4YYJD5IF","schema_version":"1.0","canonical_sha256":"e63091f505a60c003d624fc7f0c004c15e79a7cc51d1bc9c56f7162ecf55049b","source":{"kind":"arxiv","id":"2601.21406","version":3},"attestation_state":"computed","paper":{"title":"Generation Enhances Understanding in Unified Multimodal Models via Multi-Representation Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Chun Yuan, Guanhua Chen, Hongyang Wei, Kangrui Cen, Xiangxiang Chu, Yong Wang, Zihan Su","submitted_at":"2026-01-29T08:42:25Z","abstract_excerpt":"Unified Multimodal Models (UMMs) integrate both visual understanding and generation within a single framework. Their ultimate aspiration is to create a cycle where understanding and generation mutually reinforce each other. While recent post-training methods have successfully leveraged understanding to enhance generation, the reverse direction of utilizing generation to improve understanding remains largely unexplored. In this work, we propose UniMRG (Unified Multi-Representation Generation), a simple yet effective architecture-agnostic post-training method. UniMRG enhances the understanding c"},"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":"2601.21406","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-01-29T08:42:25Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"5d610ef98ee0d98d9d650435a2ed5177bc633d1bea312dd7cc0ccc1c30377482","abstract_canon_sha256":"0f0ad062253933bf65e4c49991b01c0d15511ea083528480beffc90ce9c677aa"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-26T02:04:03.861058Z","signature_b64":"9dR6Wvy6OVPAx4/fk+UIVeAGoFviZhzsRyRsfvElRYAGYEJjK9MzjKGsDA4uIpzzp1fX2HdMbi8lEoYa1A4wDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e63091f505a60c003d624fc7f0c004c15e79a7cc51d1bc9c56f7162ecf55049b","last_reissued_at":"2026-05-26T02:04:03.860153Z","signature_status":"signed_v1","first_computed_at":"2026-05-26T02:04:03.860153Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Generation Enhances Understanding in Unified Multimodal Models via Multi-Representation Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Chun Yuan, Guanhua Chen, Hongyang Wei, Kangrui Cen, Xiangxiang Chu, Yong Wang, Zihan Su","submitted_at":"2026-01-29T08:42:25Z","abstract_excerpt":"Unified Multimodal Models (UMMs) integrate both visual understanding and generation within a single framework. Their ultimate aspiration is to create a cycle where understanding and generation mutually reinforce each other. While recent post-training methods have successfully leveraged understanding to enhance generation, the reverse direction of utilizing generation to improve understanding remains largely unexplored. In this work, we propose UniMRG (Unified Multi-Representation Generation), a simple yet effective architecture-agnostic post-training method. UniMRG enhances the understanding c"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2601.21406","kind":"arxiv","version":3},"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/2601.21406/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":"2601.21406","created_at":"2026-05-26T02:04:03.860282+00:00"},{"alias_kind":"arxiv_version","alias_value":"2601.21406v3","created_at":"2026-05-26T02:04:03.860282+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.21406","created_at":"2026-05-26T02:04:03.860282+00:00"},{"alias_kind":"pith_short_12","alias_value":"4YYJD5IFUYGA","created_at":"2026-05-26T02:04:03.860282+00:00"},{"alias_kind":"pith_short_16","alias_value":"4YYJD5IFUYGAAPLC","created_at":"2026-05-26T02:04:03.860282+00:00"},{"alias_kind":"pith_short_8","alias_value":"4YYJD5IF","created_at":"2026-05-26T02:04:03.860282+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":4,"sample":[{"citing_arxiv_id":"2605.17766","citing_title":"LatentUMM: Dual Latent Alignment for Unified Multimodal Models","ref_index":37,"is_internal_anchor":true},{"citing_arxiv_id":"2605.18714","citing_title":"Semantic Generative Tuning for Unified Multimodal Models","ref_index":53,"is_internal_anchor":true},{"citing_arxiv_id":"2605.16961","citing_title":"Latent Action Control for Reasoning-Guided Unified Image Generation","ref_index":31,"is_internal_anchor":true},{"citing_arxiv_id":"2604.26341","citing_title":"SpatialFusion: Endowing Unified Image Generation with Intrinsic 3D Geometric Awareness","ref_index":46,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/4YYJD5IFUYGAAPLCJ7D7BQAEYF","json":"https://pith.science/pith/4YYJD5IFUYGAAPLCJ7D7BQAEYF.json","graph_json":"https://pith.science/api/pith-number/4YYJD5IFUYGAAPLCJ7D7BQAEYF/graph.json","events_json":"https://pith.science/api/pith-number/4YYJD5IFUYGAAPLCJ7D7BQAEYF/events.json","paper":"https://pith.science/paper/4YYJD5IF"},"agent_actions":{"view_html":"https://pith.science/pith/4YYJD5IFUYGAAPLCJ7D7BQAEYF","download_json":"https://pith.science/pith/4YYJD5IFUYGAAPLCJ7D7BQAEYF.json","view_paper":"https://pith.science/paper/4YYJD5IF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2601.21406&json=true","fetch_graph":"https://pith.science/api/pith-number/4YYJD5IFUYGAAPLCJ7D7BQAEYF/graph.json","fetch_events":"https://pith.science/api/pith-number/4YYJD5IFUYGAAPLCJ7D7BQAEYF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4YYJD5IFUYGAAPLCJ7D7BQAEYF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4YYJD5IFUYGAAPLCJ7D7BQAEYF/action/storage_attestation","attest_author":"https://pith.science/pith/4YYJD5IFUYGAAPLCJ7D7BQAEYF/action/author_attestation","sign_citation":"https://pith.science/pith/4YYJD5IFUYGAAPLCJ7D7BQAEYF/action/citation_signature","submit_replication":"https://pith.science/pith/4YYJD5IFUYGAAPLCJ7D7BQAEYF/action/replication_record"}},"created_at":"2026-05-26T02:04:03.860282+00:00","updated_at":"2026-05-26T02:04:03.860282+00:00"}