{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:AB5OYI3QWC6J5DBHFP374UCGTC","short_pith_number":"pith:AB5OYI3Q","schema_version":"1.0","canonical_sha256":"007aec2370b0bc9e8c272bf7fe504698857211fc963f2d5295a18ea4842ad671","source":{"kind":"arxiv","id":"2509.20328","version":2},"attestation_state":"computed","paper":{"title":"Video models are zero-shot learners and reasoners","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Generative video models like Veo 3 perform zero-shot object segmentation, edge detection, physics understanding, affordance recognition, tool simulation, and early visual reasoning such as maze and symmetry solving.","cross_cats":["cs.AI","cs.CV","cs.RO"],"primary_cat":"cs.LG","authors_text":"Been Kim, Kevin Swersky, Nick Matarese, Paul Vicol, Priyank Jaini, Robert Geirhos, Shixiang Shane Gu, Thadd\\\"aus Wiedemer, Yuxuan Li","submitted_at":"2025-09-24T17:17:27Z","abstract_excerpt":"The remarkable zero-shot capabilities of Large Language Models (LLMs) have propelled natural language processing from task-specific models to unified, generalist foundation models. This transformation emerged from simple primitives: large, generative models trained on web-scale data. Curiously, the same primitives apply to today's generative video models. Could video models be on a trajectory towards general-purpose vision understanding, much like LLMs developed general-purpose language understanding? We demonstrate that Veo 3 can solve a broad variety of tasks it wasn't explicitly trained for"},"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":true,"formal_links_present":true},"canonical_record":{"source":{"id":"2509.20328","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-09-24T17:17:27Z","cross_cats_sorted":["cs.AI","cs.CV","cs.RO"],"title_canon_sha256":"71d3cd4322b4002941d6b5b5741e8571148bd602b0cb7ea08d2b9a2ffff8c90e","abstract_canon_sha256":"21c914ce64dfc70a482581f938246ca5411da0dccdb7629a162004911adf78f4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T02:40:55.632905Z","signature_b64":"jLxpoau8yw7togBodVcjwoadkRSwXywr8D2V62GegU1mnHfNB0EnfxjxGYgw72bJJt0obkD1eW1A5ulK3tfrCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"007aec2370b0bc9e8c272bf7fe504698857211fc963f2d5295a18ea4842ad671","last_reissued_at":"2026-05-18T02:40:55.632229Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T02:40:55.632229Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Video models are zero-shot learners and reasoners","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Generative video models like Veo 3 perform zero-shot object segmentation, edge detection, physics understanding, affordance recognition, tool simulation, and early visual reasoning such as maze and symmetry solving.","cross_cats":["cs.AI","cs.CV","cs.RO"],"primary_cat":"cs.LG","authors_text":"Been Kim, Kevin Swersky, Nick Matarese, Paul Vicol, Priyank Jaini, Robert Geirhos, Shixiang Shane Gu, Thadd\\\"aus Wiedemer, Yuxuan Li","submitted_at":"2025-09-24T17:17:27Z","abstract_excerpt":"The remarkable zero-shot capabilities of Large Language Models (LLMs) have propelled natural language processing from task-specific models to unified, generalist foundation models. This transformation emerged from simple primitives: large, generative models trained on web-scale data. Curiously, the same primitives apply to today's generative video models. Could video models be on a trajectory towards general-purpose vision understanding, much like LLMs developed general-purpose language understanding? We demonstrate that Veo 3 can solve a broad variety of tasks it wasn't explicitly trained for"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Veo 3 can solve a broad variety of tasks it wasn't explicitly trained for: segmenting objects, detecting edges, editing images, understanding physical properties, recognizing object affordances, simulating tool use, and more. These abilities enable early forms of visual reasoning like maze and symmetry solving.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the demonstrated capabilities are genuinely zero-shot and not the result of implicit task information in the prompts, data contamination, or post-hoc selection of successful examples.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Generative video models exhibit emergent zero-shot capabilities across perception, manipulation, and basic reasoning tasks.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Generative video models like Veo 3 perform zero-shot object segmentation, edge detection, physics understanding, affordance recognition, tool simulation, and early visual reasoning such as maze and symmetry solving.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"ff0c595c84bd51c5e2400df16dae0c52e3aeab41d2f12f05e722417cba55e8d9"},"source":{"id":"2509.20328","kind":"arxiv","version":2},"verdict":{"id":"5e2c70ea-d0de-416b-a455-834dd2668077","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-14T02:11:09.410472Z","strongest_claim":"Veo 3 can solve a broad variety of tasks it wasn't explicitly trained for: segmenting objects, detecting edges, editing images, understanding physical properties, recognizing object affordances, simulating tool use, and more. These abilities enable early forms of visual reasoning like maze and symmetry solving.","one_line_summary":"Generative video models exhibit emergent zero-shot capabilities across perception, manipulation, and basic reasoning tasks.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the demonstrated capabilities are genuinely zero-shot and not the result of implicit task information in the prompts, data contamination, or post-hoc selection of successful examples.","pith_extraction_headline":"Generative video models like Veo 3 perform zero-shot object segmentation, edge detection, physics understanding, affordance recognition, tool simulation, and early visual reasoning such as maze and symmetry solving."},"references":{"count":98,"sample":[{"doi":"","year":2024,"title":"A Survey on Large Language Models for Code Generation","work_id":"7d829146-3160-44c7-9ce4-4201ed9b654c","ref_index":1,"cited_arxiv_id":"2406.00515","is_internal_anchor":true},{"doi":"","year":2025,"title":"Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities","work_id":"008df105-2fdd-45d8-857a-8e35868aecb6","ref_index":2,"cited_arxiv_id":"2507.06261","is_internal_anchor":true},{"doi":"","year":2024,"title":"Weaver: Foundation models for creative writing","work_id":"1d3d18bc-fb96-451d-b2e5-aec2c15eb603","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2023,"title":"Multilingual machine translation with large language models: Empirical results and analysis","work_id":"8aedba3a-0fa4-4db0-94b7-ed63580404df","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2024,"title":"The AI Scientist: Towards Fully Automated Open-Ended Scientific Discovery","work_id":"56b6b58d-e73a-4317-896e-36ac5f84e957","ref_index":5,"cited_arxiv_id":"2408.06292","is_internal_anchor":true}],"resolved_work":98,"snapshot_sha256":"f08f5b262f8e19d6e3a421f4ce2444e499a2bf0eed658cf2405abffc07e7ee8f","internal_anchors":15},"formal_canon":{"evidence_count":1,"snapshot_sha256":"72b87a91f7f0f708c7eea993a55ddf90800983ff1e85b7072e7aead1f82d25f4"},"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":"2509.20328","created_at":"2026-05-18T02:40:55.632344+00:00"},{"alias_kind":"arxiv_version","alias_value":"2509.20328v2","created_at":"2026-05-18T02:40:55.632344+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.20328","created_at":"2026-05-18T02:40:55.632344+00:00"},{"alias_kind":"pith_short_12","alias_value":"AB5OYI3QWC6J","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_16","alias_value":"AB5OYI3QWC6J5DBH","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_8","alias_value":"AB5OYI3Q","created_at":"2026-05-18T12:33:37.589309+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":48,"internal_anchor_count":48,"sample":[{"citing_arxiv_id":"2604.17623","citing_title":"ViPS: Video-informed Pose Spaces for Auto-Rigged Meshes","ref_index":48,"is_internal_anchor":true},{"citing_arxiv_id":"2605.22570","citing_title":"VGenST-Bench: A Benchmark for Spatio-Temporal Reasoning via Active Video Synthesis","ref_index":78,"is_internal_anchor":true},{"citing_arxiv_id":"2605.22818","citing_title":"MotiMotion: Motion-Controlled Video Generation with Visual Reasoning","ref_index":93,"is_internal_anchor":true},{"citing_arxiv_id":"2602.04876","citing_title":"PerpetualWonder: Long-Horizon Action-Conditioned 4D Scene Generation","ref_index":45,"is_internal_anchor":true},{"citing_arxiv_id":"2604.16503","citing_title":"Motif-Video 2B: Technical Report","ref_index":43,"is_internal_anchor":true},{"citing_arxiv_id":"2605.17423","citing_title":"Soap2Soap: Long Cinematic Video Remaking via Multi-Agent Collaboration","ref_index":43,"is_internal_anchor":true},{"citing_arxiv_id":"2605.17522","citing_title":"RoboFlow4D: A Lightweight Flow World Model Toward Real-Time Flow-Guided Robotic Manipulation","ref_index":29,"is_internal_anchor":true},{"citing_arxiv_id":"2605.18365","citing_title":"GeoFlow: Enforcing Implicit Geometric Consistency in Video Generation","ref_index":74,"is_internal_anchor":true},{"citing_arxiv_id":"2605.19242","citing_title":"PhyWorld: Physics-Faithful World Model for Video Generation","ref_index":3,"is_internal_anchor":true},{"citing_arxiv_id":"2605.19360","citing_title":"Scalable, Energy-Efficient Optical-Neural Architecture for Multiplexed Deepfake Video Detection","ref_index":8,"is_internal_anchor":true},{"citing_arxiv_id":"2605.15458","citing_title":"Video Models Can Reason with Verifiable Rewards","ref_index":40,"is_internal_anchor":true},{"citing_arxiv_id":"2510.10254","citing_title":"Are Video Models Emerging as Zero-Shot Learners and Reasoners in Medical Imaging?","ref_index":36,"is_internal_anchor":true},{"citing_arxiv_id":"2511.04570","citing_title":"Thinking with Video: Video Generation as a Promising Multimodal Reasoning Paradigm","ref_index":39,"is_internal_anchor":true},{"citing_arxiv_id":"2511.17792","citing_title":"Target-Bench: Can Video World Models Achieve Mapless Path Planning with Semantic Targets?","ref_index":35,"is_internal_anchor":true},{"citing_arxiv_id":"2512.07469","citing_title":"VideoCoF: Unified Video Editing with Temporal Reasoner","ref_index":36,"is_internal_anchor":true},{"citing_arxiv_id":"2512.23421","citing_title":"DriveLaW:Unifying Planning and Video Generation in a Latent Driving World","ref_index":72,"is_internal_anchor":true},{"citing_arxiv_id":"2512.04678","citing_title":"Reward Forcing: Efficient Streaming Video Generation with Rewarded Distribution Matching Distillation","ref_index":76,"is_internal_anchor":true},{"citing_arxiv_id":"2602.06949","citing_title":"DreamDojo: A Generalist Robot World Model from Large-Scale Human Videos","ref_index":100,"is_internal_anchor":true},{"citing_arxiv_id":"2601.08565","citing_title":"Rewriting Video: Text-Driven Reauthoring of Video Footage","ref_index":24,"is_internal_anchor":true},{"citing_arxiv_id":"2512.16776","citing_title":"Kling-Omni Technical Report","ref_index":33,"is_internal_anchor":true},{"citing_arxiv_id":"2512.14614","citing_title":"WorldPlay: Towards Long-Term Geometric Consistency for Real-Time Interactive World Modeling","ref_index":65,"is_internal_anchor":true},{"citing_arxiv_id":"2512.15692","citing_title":"mimic-video: Video-Action Models for Generalizable Robot Control Beyond VLAs","ref_index":57,"is_internal_anchor":true},{"citing_arxiv_id":"2603.17834","citing_title":"Generative Control as Optimization: Time Unconditional Flow Matching for Adaptive and Robust Robotic Control","ref_index":28,"is_internal_anchor":true},{"citing_arxiv_id":"2604.20329","citing_title":"Image Generators are Generalist Vision Learners","ref_index":27,"is_internal_anchor":true},{"citing_arxiv_id":"2603.21210","citing_title":"Pretrained Video Models as Differentiable Physics Simulators for Urban Wind Flows","ref_index":56,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":1,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/AB5OYI3QWC6J5DBHFP374UCGTC","json":"https://pith.science/pith/AB5OYI3QWC6J5DBHFP374UCGTC.json","graph_json":"https://pith.science/api/pith-number/AB5OYI3QWC6J5DBHFP374UCGTC/graph.json","events_json":"https://pith.science/api/pith-number/AB5OYI3QWC6J5DBHFP374UCGTC/events.json","paper":"https://pith.science/paper/AB5OYI3Q"},"agent_actions":{"view_html":"https://pith.science/pith/AB5OYI3QWC6J5DBHFP374UCGTC","download_json":"https://pith.science/pith/AB5OYI3QWC6J5DBHFP374UCGTC.json","view_paper":"https://pith.science/paper/AB5OYI3Q","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2509.20328&json=true","fetch_graph":"https://pith.science/api/pith-number/AB5OYI3QWC6J5DBHFP374UCGTC/graph.json","fetch_events":"https://pith.science/api/pith-number/AB5OYI3QWC6J5DBHFP374UCGTC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/AB5OYI3QWC6J5DBHFP374UCGTC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/AB5OYI3QWC6J5DBHFP374UCGTC/action/storage_attestation","attest_author":"https://pith.science/pith/AB5OYI3QWC6J5DBHFP374UCGTC/action/author_attestation","sign_citation":"https://pith.science/pith/AB5OYI3QWC6J5DBHFP374UCGTC/action/citation_signature","submit_replication":"https://pith.science/pith/AB5OYI3QWC6J5DBHFP374UCGTC/action/replication_record"}},"created_at":"2026-05-18T02:40:55.632344+00:00","updated_at":"2026-05-18T02:40:55.632344+00:00"}