{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:JURFBPZVF2RJ6D56IPDIQNS4NL","short_pith_number":"pith:JURFBPZV","schema_version":"1.0","canonical_sha256":"4d2250bf352ea29f0fbe43c688365c6afad94abe17d8416b465f719aafcf69c8","source":{"kind":"arxiv","id":"2502.15803","version":1},"attestation_state":"computed","paper":{"title":"Megrez-Omni Technical Report","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Boxun Li, Congyi Liu, Dong Zhou, Guohao Dai, Guowei Niu, Haiyang Xu, Shengen Yan, Tao Yuan, Weilin Liu, Yadong Li, Yueqing Zhuang, Yu Wang, Zheyue Tan, Zhiyuan Li, Zhuyu Yao","submitted_at":"2025-02-19T06:14:14Z","abstract_excerpt":"In this work, we present the Megrez models, comprising a language model (Megrez-3B-Instruct) and a multimodal model (Megrez-3B-Omni). These models are designed to deliver fast inference, compactness, and robust edge-side intelligence through a software-hardware co-design approach. Megrez-3B-Instruct offers several advantages, including high accuracy, high speed, ease of use, and a wide range of applications. Building on Megrez-3B-Instruct, Megrez-3B-Omni is an on-device multimodal understanding LLM that supports image, text, and audio analysis. It achieves state-of-the-art accuracy across all "},"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":"2502.15803","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-19T06:14:14Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"ed02cb062eddf6c81882bab4a8f17a61daabddd161c2ce6c87769d9f10dc8601","abstract_canon_sha256":"1338ddc48f1439a8bc8fcab1a95c78171e2fb33a06081fd3b65afecc78c266ac"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:20:19.684356Z","signature_b64":"kP266MEb2CmAbdPZwctWvY9Y2MC3ctnc+dm9sAsYE89GArNYkV9Uf92aThTOSzf744PT4qMcaaTul49lPzD8BA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4d2250bf352ea29f0fbe43c688365c6afad94abe17d8416b465f719aafcf69c8","last_reissued_at":"2026-07-05T10:20:19.683742Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:20:19.683742Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Megrez-Omni Technical Report","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Boxun Li, Congyi Liu, Dong Zhou, Guohao Dai, Guowei Niu, Haiyang Xu, Shengen Yan, Tao Yuan, Weilin Liu, Yadong Li, Yueqing Zhuang, Yu Wang, Zheyue Tan, Zhiyuan Li, Zhuyu Yao","submitted_at":"2025-02-19T06:14:14Z","abstract_excerpt":"In this work, we present the Megrez models, comprising a language model (Megrez-3B-Instruct) and a multimodal model (Megrez-3B-Omni). These models are designed to deliver fast inference, compactness, and robust edge-side intelligence through a software-hardware co-design approach. Megrez-3B-Instruct offers several advantages, including high accuracy, high speed, ease of use, and a wide range of applications. Building on Megrez-3B-Instruct, Megrez-3B-Omni is an on-device multimodal understanding LLM that supports image, text, and audio analysis. It achieves state-of-the-art accuracy across all "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.15803","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/2502.15803/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":"2502.15803","created_at":"2026-07-05T10:20:19.683817+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.15803v1","created_at":"2026-07-05T10:20:19.683817+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.15803","created_at":"2026-07-05T10:20:19.683817+00:00"},{"alias_kind":"pith_short_12","alias_value":"JURFBPZVF2RJ","created_at":"2026-07-05T10:20:19.683817+00:00"},{"alias_kind":"pith_short_16","alias_value":"JURFBPZVF2RJ6D56","created_at":"2026-07-05T10:20:19.683817+00:00"},{"alias_kind":"pith_short_8","alias_value":"JURFBPZV","created_at":"2026-07-05T10:20:19.683817+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.07433","citing_title":"Watch, Remember, Reason: Human-View Video Understanding with MLLMs","ref_index":125,"is_internal_anchor":false},{"citing_arxiv_id":"2605.01278","citing_title":"Valley3: Scaling Omni Foundation Models for E-commerce","ref_index":38,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JURFBPZVF2RJ6D56IPDIQNS4NL","json":"https://pith.science/pith/JURFBPZVF2RJ6D56IPDIQNS4NL.json","graph_json":"https://pith.science/api/pith-number/JURFBPZVF2RJ6D56IPDIQNS4NL/graph.json","events_json":"https://pith.science/api/pith-number/JURFBPZVF2RJ6D56IPDIQNS4NL/events.json","paper":"https://pith.science/paper/JURFBPZV"},"agent_actions":{"view_html":"https://pith.science/pith/JURFBPZVF2RJ6D56IPDIQNS4NL","download_json":"https://pith.science/pith/JURFBPZVF2RJ6D56IPDIQNS4NL.json","view_paper":"https://pith.science/paper/JURFBPZV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.15803&json=true","fetch_graph":"https://pith.science/api/pith-number/JURFBPZVF2RJ6D56IPDIQNS4NL/graph.json","fetch_events":"https://pith.science/api/pith-number/JURFBPZVF2RJ6D56IPDIQNS4NL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JURFBPZVF2RJ6D56IPDIQNS4NL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JURFBPZVF2RJ6D56IPDIQNS4NL/action/storage_attestation","attest_author":"https://pith.science/pith/JURFBPZVF2RJ6D56IPDIQNS4NL/action/author_attestation","sign_citation":"https://pith.science/pith/JURFBPZVF2RJ6D56IPDIQNS4NL/action/citation_signature","submit_replication":"https://pith.science/pith/JURFBPZVF2RJ6D56IPDIQNS4NL/action/replication_record"}},"created_at":"2026-07-05T10:20:19.683817+00:00","updated_at":"2026-07-05T10:20:19.683817+00:00"}