{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:PIZHHPHORGFEETEGEWLLIMEVME","short_pith_number":"pith:PIZHHPHO","schema_version":"1.0","canonical_sha256":"7a3273bcee898a424c862596b4309561249e72deb4c5565ffd228828802487ce","source":{"kind":"arxiv","id":"2508.12109","version":1},"attestation_state":"computed","paper":{"title":"Simple o3: Towards Interleaved Vision-Language Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Qianglong Chen, Shijie Guo, Siyuan Wang, Ye Wang, Zejun Li, Zhirui Zhang, Zhongyu Wei","submitted_at":"2025-08-16T17:15:39Z","abstract_excerpt":"Multimodal Large Language Models (MLLMs) have shown impressive performance on vision-language tasks, but their long Chain-of-Thought (CoT) capabilities in multimodal scenarios remain underexplored. Inspired by OpenAI's o3 model, which emulates human-like ''thinking with image'' through iterative visual transformations and linguistic reasoning, we propose Simple o3, an end-to-end framework that integrates dynamic tool interactions (e.g., cropping, zooming, and reusing) into interleaved vision-language reasoning via supervised fine-tuning (SFT). Our approach features a scalable data synthesis pi"},"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":"2508.12109","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-08-16T17:15:39Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"22cc2be6349b176ca112b92cdd8725cc56f6a5b76fe10a14d8b581cdc03f7b72","abstract_canon_sha256":"98c000df40625c5b76bc9c702da9698c984cb30b0014dad8a942af0605ab5132"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:55:00.245705Z","signature_b64":"Og7Y89/YhkU6dCzJ+kNuTt8mohxqyFdfymrdd1BD7YwyPdlpWWOudv0Y/gyRkxiIEptX0gjnmC4a+9+RdrTdAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7a3273bcee898a424c862596b4309561249e72deb4c5565ffd228828802487ce","last_reissued_at":"2026-07-05T11:55:00.245232Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:55:00.245232Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Simple o3: Towards Interleaved Vision-Language Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Qianglong Chen, Shijie Guo, Siyuan Wang, Ye Wang, Zejun Li, Zhirui Zhang, Zhongyu Wei","submitted_at":"2025-08-16T17:15:39Z","abstract_excerpt":"Multimodal Large Language Models (MLLMs) have shown impressive performance on vision-language tasks, but their long Chain-of-Thought (CoT) capabilities in multimodal scenarios remain underexplored. Inspired by OpenAI's o3 model, which emulates human-like ''thinking with image'' through iterative visual transformations and linguistic reasoning, we propose Simple o3, an end-to-end framework that integrates dynamic tool interactions (e.g., cropping, zooming, and reusing) into interleaved vision-language reasoning via supervised fine-tuning (SFT). Our approach features a scalable data synthesis pi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.12109","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/2508.12109/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":"2508.12109","created_at":"2026-07-05T11:55:00.245290+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.12109v1","created_at":"2026-07-05T11:55:00.245290+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.12109","created_at":"2026-07-05T11:55:00.245290+00:00"},{"alias_kind":"pith_short_12","alias_value":"PIZHHPHORGFE","created_at":"2026-07-05T11:55:00.245290+00:00"},{"alias_kind":"pith_short_16","alias_value":"PIZHHPHORGFEETEG","created_at":"2026-07-05T11:55:00.245290+00:00"},{"alias_kind":"pith_short_8","alias_value":"PIZHHPHO","created_at":"2026-07-05T11:55:00.245290+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.00096","citing_title":"Diversity Over Frequency: Rethinking Tool Use in Visual Chain-of-Thought Agents","ref_index":62,"is_internal_anchor":false},{"citing_arxiv_id":"2605.27959","citing_title":"ROVER: Routing Object-Centric Visual Evidence for Grounded Multi-Image Reasoning","ref_index":61,"is_internal_anchor":false},{"citing_arxiv_id":"2512.08980","citing_title":"Training Multi-Image Vision Agents via End2End Reinforcement Learning","ref_index":43,"is_internal_anchor":false},{"citing_arxiv_id":"2601.15724","citing_title":"VideoThinker: Building Agentic VideoLLMs with LLM-Guided Tool Reasoning","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2604.19945","citing_title":"Visual Reasoning through Tool-supervised Reinforcement Learning","ref_index":26,"is_internal_anchor":false},{"citing_arxiv_id":"2604.09167","citing_title":"MAG-3D: Multi-Agent Grounded Reasoning for 3D Understanding","ref_index":43,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PIZHHPHORGFEETEGEWLLIMEVME","json":"https://pith.science/pith/PIZHHPHORGFEETEGEWLLIMEVME.json","graph_json":"https://pith.science/api/pith-number/PIZHHPHORGFEETEGEWLLIMEVME/graph.json","events_json":"https://pith.science/api/pith-number/PIZHHPHORGFEETEGEWLLIMEVME/events.json","paper":"https://pith.science/paper/PIZHHPHO"},"agent_actions":{"view_html":"https://pith.science/pith/PIZHHPHORGFEETEGEWLLIMEVME","download_json":"https://pith.science/pith/PIZHHPHORGFEETEGEWLLIMEVME.json","view_paper":"https://pith.science/paper/PIZHHPHO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.12109&json=true","fetch_graph":"https://pith.science/api/pith-number/PIZHHPHORGFEETEGEWLLIMEVME/graph.json","fetch_events":"https://pith.science/api/pith-number/PIZHHPHORGFEETEGEWLLIMEVME/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PIZHHPHORGFEETEGEWLLIMEVME/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PIZHHPHORGFEETEGEWLLIMEVME/action/storage_attestation","attest_author":"https://pith.science/pith/PIZHHPHORGFEETEGEWLLIMEVME/action/author_attestation","sign_citation":"https://pith.science/pith/PIZHHPHORGFEETEGEWLLIMEVME/action/citation_signature","submit_replication":"https://pith.science/pith/PIZHHPHORGFEETEGEWLLIMEVME/action/replication_record"}},"created_at":"2026-07-05T11:55:00.245290+00:00","updated_at":"2026-07-05T11:55:00.245290+00:00"}