{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:DXBGCOOXMYILH45AGG6N32VETH","short_pith_number":"pith:DXBGCOOX","schema_version":"1.0","canonical_sha256":"1dc26139d76610b3f3a031bcddeaa499f6ee22da2993a834968a4454a0818021","source":{"kind":"arxiv","id":"2505.23727","version":1},"attestation_state":"computed","paper":{"title":"PixelThink: Towards Efficient Chain-of-Pixel Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.MM"],"primary_cat":"cs.CV","authors_text":"Gongfan Fang, Jianke Zhu, Jianyun Xu, Lingdong Kong, Qiang Li, Sheng Yang, Song Wang, Xiangtai Li, Xinchao Wang","submitted_at":"2025-05-29T17:55:49Z","abstract_excerpt":"Existing reasoning segmentation approaches typically fine-tune multimodal large language models (MLLMs) using image-text pairs and corresponding mask labels. However, they exhibit limited generalization to out-of-distribution scenarios without an explicit reasoning process. Although recent efforts leverage reinforcement learning through group-relative policy optimization (GRPO) to enhance reasoning ability, they often suffer from overthinking - producing uniformly verbose reasoning chains irrespective of task complexity. This results in elevated computational costs and limited control over rea"},"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":"2505.23727","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-05-29T17:55:49Z","cross_cats_sorted":["cs.MM"],"title_canon_sha256":"ce68598744f10edeb29454158273f47f3ed2549e31c7a81190c718123e9c0028","abstract_canon_sha256":"b1e014a0a26fae27e8d31589f155de1a8a7913d2a6cca367f8102808385c08ce"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:12:13.339253Z","signature_b64":"jCzQdl4ND0uemELK4k91O+l6bxyqk9JQ0I9KoRImCUPn8XMSGv3joWkW3ssCVODsHV1Xo+CO9f9wtmoM31FeBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1dc26139d76610b3f3a031bcddeaa499f6ee22da2993a834968a4454a0818021","last_reissued_at":"2026-07-05T11:12:13.338773Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:12:13.338773Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"PixelThink: Towards Efficient Chain-of-Pixel Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.MM"],"primary_cat":"cs.CV","authors_text":"Gongfan Fang, Jianke Zhu, Jianyun Xu, Lingdong Kong, Qiang Li, Sheng Yang, Song Wang, Xiangtai Li, Xinchao Wang","submitted_at":"2025-05-29T17:55:49Z","abstract_excerpt":"Existing reasoning segmentation approaches typically fine-tune multimodal large language models (MLLMs) using image-text pairs and corresponding mask labels. However, they exhibit limited generalization to out-of-distribution scenarios without an explicit reasoning process. Although recent efforts leverage reinforcement learning through group-relative policy optimization (GRPO) to enhance reasoning ability, they often suffer from overthinking - producing uniformly verbose reasoning chains irrespective of task complexity. This results in elevated computational costs and limited control over rea"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.23727","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/2505.23727/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":"2505.23727","created_at":"2026-07-05T11:12:13.338832+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.23727v1","created_at":"2026-07-05T11:12:13.338832+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.23727","created_at":"2026-07-05T11:12:13.338832+00:00"},{"alias_kind":"pith_short_12","alias_value":"DXBGCOOXMYIL","created_at":"2026-07-05T11:12:13.338832+00:00"},{"alias_kind":"pith_short_16","alias_value":"DXBGCOOXMYILH45A","created_at":"2026-07-05T11:12:13.338832+00:00"},{"alias_kind":"pith_short_8","alias_value":"DXBGCOOX","created_at":"2026-07-05T11:12:13.338832+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":183,"is_internal_anchor":false},{"citing_arxiv_id":"2607.02497","citing_title":"Seek to Segment: Active Perception for Panoramic Referring Segmentation","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2604.11789","citing_title":"LMMs Meet Object-Centric Vision: Understanding, Segmentation, Editing and Generation","ref_index":167,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07334","citing_title":"RCoT-Seg: Reinforced Chain-of-Thought for Video Reasoning and Segmentation","ref_index":23,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DXBGCOOXMYILH45AGG6N32VETH","json":"https://pith.science/pith/DXBGCOOXMYILH45AGG6N32VETH.json","graph_json":"https://pith.science/api/pith-number/DXBGCOOXMYILH45AGG6N32VETH/graph.json","events_json":"https://pith.science/api/pith-number/DXBGCOOXMYILH45AGG6N32VETH/events.json","paper":"https://pith.science/paper/DXBGCOOX"},"agent_actions":{"view_html":"https://pith.science/pith/DXBGCOOXMYILH45AGG6N32VETH","download_json":"https://pith.science/pith/DXBGCOOXMYILH45AGG6N32VETH.json","view_paper":"https://pith.science/paper/DXBGCOOX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.23727&json=true","fetch_graph":"https://pith.science/api/pith-number/DXBGCOOXMYILH45AGG6N32VETH/graph.json","fetch_events":"https://pith.science/api/pith-number/DXBGCOOXMYILH45AGG6N32VETH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DXBGCOOXMYILH45AGG6N32VETH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DXBGCOOXMYILH45AGG6N32VETH/action/storage_attestation","attest_author":"https://pith.science/pith/DXBGCOOXMYILH45AGG6N32VETH/action/author_attestation","sign_citation":"https://pith.science/pith/DXBGCOOXMYILH45AGG6N32VETH/action/citation_signature","submit_replication":"https://pith.science/pith/DXBGCOOXMYILH45AGG6N32VETH/action/replication_record"}},"created_at":"2026-07-05T11:12:13.338832+00:00","updated_at":"2026-07-05T11:12:13.338832+00:00"}