{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:TQMR47DMULMJOV577UAKUYZ3GW","short_pith_number":"pith:TQMR47DM","schema_version":"1.0","canonical_sha256":"9c191e7c6ca2d89757bffd00aa633b35a07e9701d6536cc6e370e9a58a2dc9f8","source":{"kind":"arxiv","id":"2506.22434","version":1},"attestation_state":"computed","paper":{"title":"MiCo: Multi-image Contrast for Reinforcement Visual Reasoning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Hengshuang Zhao, Mingkang Zhu, Shaoteng Liu, Xiang Bai, Xiaogang Xu, Xiaoyang Wu, Xi Chen, Yu Liu","submitted_at":"2025-06-27T17:59:27Z","abstract_excerpt":"This work explores enabling Chain-of-Thought (CoT) reasoning to link visual cues across multiple images. A straightforward solution is to adapt rule-based reinforcement learning for Vision-Language Models (VLMs). However, such methods typically rely on manually curated question-answer pairs, which can be particularly challenging when dealing with fine grained visual details and complex logic across images. Inspired by self-supervised visual representation learning, we observe that images contain inherent constraints that can serve as supervision. Based on this insight, we construct image tripl"},"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":"2506.22434","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-06-27T17:59:27Z","cross_cats_sorted":[],"title_canon_sha256":"828b0a93516a4615aef040894aeb6e720723e720190ce13aa22f2c8b50fa4a1c","abstract_canon_sha256":"44cfa1a9f2c3a37a2ffb7c3329bffbbef51d7b427d83641ca3a0d8fac38e7212"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:28:27.450434Z","signature_b64":"kUtLP1/kuWYOrBPh1bqera/tlSvglmjrY9Oe5k209YtNeyse0Gxydsb4aBz2RVD2SinrxmGuesTxFLXWnOW6DA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9c191e7c6ca2d89757bffd00aa633b35a07e9701d6536cc6e370e9a58a2dc9f8","last_reissued_at":"2026-07-05T11:28:27.449911Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:28:27.449911Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MiCo: Multi-image Contrast for Reinforcement Visual Reasoning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Hengshuang Zhao, Mingkang Zhu, Shaoteng Liu, Xiang Bai, Xiaogang Xu, Xiaoyang Wu, Xi Chen, Yu Liu","submitted_at":"2025-06-27T17:59:27Z","abstract_excerpt":"This work explores enabling Chain-of-Thought (CoT) reasoning to link visual cues across multiple images. A straightforward solution is to adapt rule-based reinforcement learning for Vision-Language Models (VLMs). However, such methods typically rely on manually curated question-answer pairs, which can be particularly challenging when dealing with fine grained visual details and complex logic across images. Inspired by self-supervised visual representation learning, we observe that images contain inherent constraints that can serve as supervision. Based on this insight, we construct image tripl"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.22434","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/2506.22434/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":"2506.22434","created_at":"2026-07-05T11:28:27.449977+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.22434v1","created_at":"2026-07-05T11:28:27.449977+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.22434","created_at":"2026-07-05T11:28:27.449977+00:00"},{"alias_kind":"pith_short_12","alias_value":"TQMR47DMULMJ","created_at":"2026-07-05T11:28:27.449977+00:00"},{"alias_kind":"pith_short_16","alias_value":"TQMR47DMULMJOV57","created_at":"2026-07-05T11:28:27.449977+00:00"},{"alias_kind":"pith_short_8","alias_value":"TQMR47DM","created_at":"2026-07-05T11:28:27.449977+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.26196","citing_title":"From Structure to Synergy: A Survey of Vision-Language Perception Paradigm Evolution in Multimodal Large Language Models","ref_index":193,"is_internal_anchor":false},{"citing_arxiv_id":"2606.21337","citing_title":"DataClaw0: Agentic Tailoring Multimodal Data from Raw Streams","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2606.28266","citing_title":"RSICCLLM: A Multimodal Large Language Model for Remote Sensing Image Change Captioning","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2604.22498","citing_title":"CGC: Compositional Grounded Contrast for Fine-Grained Multi-Image Understanding","ref_index":13,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/TQMR47DMULMJOV577UAKUYZ3GW","json":"https://pith.science/pith/TQMR47DMULMJOV577UAKUYZ3GW.json","graph_json":"https://pith.science/api/pith-number/TQMR47DMULMJOV577UAKUYZ3GW/graph.json","events_json":"https://pith.science/api/pith-number/TQMR47DMULMJOV577UAKUYZ3GW/events.json","paper":"https://pith.science/paper/TQMR47DM"},"agent_actions":{"view_html":"https://pith.science/pith/TQMR47DMULMJOV577UAKUYZ3GW","download_json":"https://pith.science/pith/TQMR47DMULMJOV577UAKUYZ3GW.json","view_paper":"https://pith.science/paper/TQMR47DM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.22434&json=true","fetch_graph":"https://pith.science/api/pith-number/TQMR47DMULMJOV577UAKUYZ3GW/graph.json","fetch_events":"https://pith.science/api/pith-number/TQMR47DMULMJOV577UAKUYZ3GW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TQMR47DMULMJOV577UAKUYZ3GW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TQMR47DMULMJOV577UAKUYZ3GW/action/storage_attestation","attest_author":"https://pith.science/pith/TQMR47DMULMJOV577UAKUYZ3GW/action/author_attestation","sign_citation":"https://pith.science/pith/TQMR47DMULMJOV577UAKUYZ3GW/action/citation_signature","submit_replication":"https://pith.science/pith/TQMR47DMULMJOV577UAKUYZ3GW/action/replication_record"}},"created_at":"2026-07-05T11:28:27.449977+00:00","updated_at":"2026-07-05T11:28:27.449977+00:00"}