{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:NEJSSQX73EMZUT4GQRG76RRWLV","short_pith_number":"pith:NEJSSQX7","schema_version":"1.0","canonical_sha256":"69132942ffd9199a4f86844dff46365d76fc71d2708d2db49152a1991144322f","source":{"kind":"arxiv","id":"2507.07424","version":1},"attestation_state":"computed","paper":{"title":"Corvid: Improving Multimodal Large Language Models Towards Chain-of-Thought Reasoning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chao Ma, Hanwang Zhang, Jingjing Jiang, Jun Luo, Xurui Song","submitted_at":"2025-07-10T04:31:56Z","abstract_excerpt":"Recent advancements in multimodal large language models (MLLMs) have demonstrated exceptional performance in multimodal perception and understanding. However, leading open-source MLLMs exhibit significant limitations in complex and structured reasoning, particularly in tasks requiring deep reasoning for decision-making and problem-solving. In this work, we present Corvid, an MLLM with enhanced chain-of-thought (CoT) reasoning capabilities. Architecturally, Corvid incorporates a hybrid vision encoder for informative visual representation and a meticulously designed connector (GateMixer) to faci"},"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":"2507.07424","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-07-10T04:31:56Z","cross_cats_sorted":[],"title_canon_sha256":"a47e565760f6a23a33e1f2474913d7fad8bc80f790bb5dc44283360a1910c630","abstract_canon_sha256":"28226107f6dd8aa4117957d7af16f70c5a19101caf5fcc788e7b3f85903c86a8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:34:57.712163Z","signature_b64":"N4Yprd5OYVMdVFSXeeiVZUEpy32BVbjBfm4LTnKXciitVRBTvkue8rqJQrK3QBdC0qNMrrzKDKGCUQCtSNrzCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"69132942ffd9199a4f86844dff46365d76fc71d2708d2db49152a1991144322f","last_reissued_at":"2026-07-05T11:34:57.711683Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:34:57.711683Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Corvid: Improving Multimodal Large Language Models Towards Chain-of-Thought Reasoning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chao Ma, Hanwang Zhang, Jingjing Jiang, Jun Luo, Xurui Song","submitted_at":"2025-07-10T04:31:56Z","abstract_excerpt":"Recent advancements in multimodal large language models (MLLMs) have demonstrated exceptional performance in multimodal perception and understanding. However, leading open-source MLLMs exhibit significant limitations in complex and structured reasoning, particularly in tasks requiring deep reasoning for decision-making and problem-solving. In this work, we present Corvid, an MLLM with enhanced chain-of-thought (CoT) reasoning capabilities. Architecturally, Corvid incorporates a hybrid vision encoder for informative visual representation and a meticulously designed connector (GateMixer) to faci"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.07424","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/2507.07424/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":"2507.07424","created_at":"2026-07-05T11:34:57.711746+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.07424v1","created_at":"2026-07-05T11:34:57.711746+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.07424","created_at":"2026-07-05T11:34:57.711746+00:00"},{"alias_kind":"pith_short_12","alias_value":"NEJSSQX73EMZ","created_at":"2026-07-05T11:34:57.711746+00:00"},{"alias_kind":"pith_short_16","alias_value":"NEJSSQX73EMZUT4G","created_at":"2026-07-05T11:34:57.711746+00:00"},{"alias_kind":"pith_short_8","alias_value":"NEJSSQX7","created_at":"2026-07-05T11:34:57.711746+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2508.06226","citing_title":"GeoLaux: A Benchmark for Evaluating MLLMs' Geometry Performance on Long-Step Problems Requiring Auxiliary Lines","ref_index":15,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NEJSSQX73EMZUT4GQRG76RRWLV","json":"https://pith.science/pith/NEJSSQX73EMZUT4GQRG76RRWLV.json","graph_json":"https://pith.science/api/pith-number/NEJSSQX73EMZUT4GQRG76RRWLV/graph.json","events_json":"https://pith.science/api/pith-number/NEJSSQX73EMZUT4GQRG76RRWLV/events.json","paper":"https://pith.science/paper/NEJSSQX7"},"agent_actions":{"view_html":"https://pith.science/pith/NEJSSQX73EMZUT4GQRG76RRWLV","download_json":"https://pith.science/pith/NEJSSQX73EMZUT4GQRG76RRWLV.json","view_paper":"https://pith.science/paper/NEJSSQX7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.07424&json=true","fetch_graph":"https://pith.science/api/pith-number/NEJSSQX73EMZUT4GQRG76RRWLV/graph.json","fetch_events":"https://pith.science/api/pith-number/NEJSSQX73EMZUT4GQRG76RRWLV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NEJSSQX73EMZUT4GQRG76RRWLV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NEJSSQX73EMZUT4GQRG76RRWLV/action/storage_attestation","attest_author":"https://pith.science/pith/NEJSSQX73EMZUT4GQRG76RRWLV/action/author_attestation","sign_citation":"https://pith.science/pith/NEJSSQX73EMZUT4GQRG76RRWLV/action/citation_signature","submit_replication":"https://pith.science/pith/NEJSSQX73EMZUT4GQRG76RRWLV/action/replication_record"}},"created_at":"2026-07-05T11:34:57.711746+00:00","updated_at":"2026-07-05T11:34:57.711746+00:00"}