{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:XJFF3DD5E426NKFESODBUYCL57","short_pith_number":"pith:XJFF3DD5","schema_version":"1.0","canonical_sha256":"ba4a5d8c7d2735e6a8a493861a604beff4d248706e98a2111acdafd545278aee","source":{"kind":"arxiv","id":"2506.23940","version":2},"attestation_state":"computed","paper":{"title":"Graft: Integrating the Domain Knowledge via Efficient Parameter Synergy for MLLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Hongyang He, Hongzhe Huang, Jianxiang An, Siliang Tang, Tianwei Lin, Wenqiao Zhang, Yang Dai, Yueting Zhuang, Zheqi Lv","submitted_at":"2025-06-30T15:07:41Z","abstract_excerpt":"Multimodal Large Language Models (MLLMs) have achieved success across various domains. However, their applicability tends to degrade when confronted with different types of data inputs, especially for MLLMs that have been fine-tuned for specific tasks. Despite its importance, the study of knowledge sharing among domain-specific MLLMs--such as those trained for mathematics or code--remains largely underexplored. To address the fragmentation of knowledge across domain-specialized MLLMs, we propose a unified parameter integration framework that enables modular composition of expert capabilities. "},"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":"2506.23940","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-30T15:07:41Z","cross_cats_sorted":[],"title_canon_sha256":"87091b49fadf74bffeb66b394f9690a98c5c8015d4ab91da5ca049fdd1790448","abstract_canon_sha256":"f5af0ef4f224f2173831ace039d80307435401a5091bcd170660b6f0f95278b2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:30:04.161826Z","signature_b64":"2Ua7gMpuPD0MxDpbvMykF7ugPPLgux6DUb2qotphxR6P/wZs/tNNuYtH26y4AtSjZ1nEIixghdps8oh9FqFOBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ba4a5d8c7d2735e6a8a493861a604beff4d248706e98a2111acdafd545278aee","last_reissued_at":"2026-07-05T11:30:04.161287Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:30:04.161287Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Graft: Integrating the Domain Knowledge via Efficient Parameter Synergy for MLLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Hongyang He, Hongzhe Huang, Jianxiang An, Siliang Tang, Tianwei Lin, Wenqiao Zhang, Yang Dai, Yueting Zhuang, Zheqi Lv","submitted_at":"2025-06-30T15:07:41Z","abstract_excerpt":"Multimodal Large Language Models (MLLMs) have achieved success across various domains. However, their applicability tends to degrade when confronted with different types of data inputs, especially for MLLMs that have been fine-tuned for specific tasks. Despite its importance, the study of knowledge sharing among domain-specific MLLMs--such as those trained for mathematics or code--remains largely underexplored. To address the fragmentation of knowledge across domain-specialized MLLMs, we propose a unified parameter integration framework that enables modular composition of expert capabilities. "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.23940","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/2506.23940/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":"2506.23940","created_at":"2026-07-05T11:30:04.161358+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.23940v2","created_at":"2026-07-05T11:30:04.161358+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.23940","created_at":"2026-07-05T11:30:04.161358+00:00"},{"alias_kind":"pith_short_12","alias_value":"XJFF3DD5E426","created_at":"2026-07-05T11:30:04.161358+00:00"},{"alias_kind":"pith_short_16","alias_value":"XJFF3DD5E426NKFE","created_at":"2026-07-05T11:30:04.161358+00:00"},{"alias_kind":"pith_short_8","alias_value":"XJFF3DD5","created_at":"2026-07-05T11:30:04.161358+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.19559","citing_title":"EgoCoT-Bench: Benchmarking Grounded and Verifiable Operation-Centric Chain of Thought Reasoning for MLLMs","ref_index":10,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/XJFF3DD5E426NKFESODBUYCL57","json":"https://pith.science/pith/XJFF3DD5E426NKFESODBUYCL57.json","graph_json":"https://pith.science/api/pith-number/XJFF3DD5E426NKFESODBUYCL57/graph.json","events_json":"https://pith.science/api/pith-number/XJFF3DD5E426NKFESODBUYCL57/events.json","paper":"https://pith.science/paper/XJFF3DD5"},"agent_actions":{"view_html":"https://pith.science/pith/XJFF3DD5E426NKFESODBUYCL57","download_json":"https://pith.science/pith/XJFF3DD5E426NKFESODBUYCL57.json","view_paper":"https://pith.science/paper/XJFF3DD5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.23940&json=true","fetch_graph":"https://pith.science/api/pith-number/XJFF3DD5E426NKFESODBUYCL57/graph.json","fetch_events":"https://pith.science/api/pith-number/XJFF3DD5E426NKFESODBUYCL57/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XJFF3DD5E426NKFESODBUYCL57/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XJFF3DD5E426NKFESODBUYCL57/action/storage_attestation","attest_author":"https://pith.science/pith/XJFF3DD5E426NKFESODBUYCL57/action/author_attestation","sign_citation":"https://pith.science/pith/XJFF3DD5E426NKFESODBUYCL57/action/citation_signature","submit_replication":"https://pith.science/pith/XJFF3DD5E426NKFESODBUYCL57/action/replication_record"}},"created_at":"2026-07-05T11:30:04.161358+00:00","updated_at":"2026-07-05T11:30:04.161358+00:00"}