{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:EJIVFRXU6P2MSHA7YTIQRPXXRC","short_pith_number":"pith:EJIVFRXU","schema_version":"1.0","canonical_sha256":"225152c6f4f3f4c91c1fc4d108bef788a12aea3102ff5ebaa7da712a657626ac","source":{"kind":"arxiv","id":"2303.08128","version":1},"attestation_state":"computed","paper":{"title":"ViperGPT: Visual Inference via Python Execution for Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"ViperGPT uses code generation to create Python programs that combine vision models for answering complex visual queries without training.","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Carl Vondrick, D\\'idac Sur\\'is, Sachit Menon","submitted_at":"2023-03-14T17:57:47Z","abstract_excerpt":"Answering visual queries is a complex task that requires both visual processing and reasoning. End-to-end models, the dominant approach for this task, do not explicitly differentiate between the two, limiting interpretability and generalization. Learning modular programs presents a promising alternative, but has proven challenging due to the difficulty of learning both the programs and modules simultaneously. We introduce ViperGPT, a framework that leverages code-generation models to compose vision-and-language models into subroutines to produce a result for any query. ViperGPT utilizes a prov"},"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":"2303.08128","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2023-03-14T17:57:47Z","cross_cats_sorted":[],"title_canon_sha256":"898de40b5c3a0c4369f3ad602ee1292bf94bf954d63805c955271fb60fc783da","abstract_canon_sha256":"c2e103dccd0a0f4cdf6a0868e0ac08d507e2a78eaf6cae4a8431e4e07a4e19f3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:38:13.427971Z","signature_b64":"6UPTUj7b+4qmbA5QdZqLcVVAgxhR0tUYGO2LeeC2o8DwvBkHAA0gfsl+XYYTprCSyuo5KvoBw+F43iKP+GgbAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"225152c6f4f3f4c91c1fc4d108bef788a12aea3102ff5ebaa7da712a657626ac","last_reissued_at":"2026-05-17T23:38:13.427145Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:38:13.427145Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ViperGPT: Visual Inference via Python Execution for Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"ViperGPT uses code generation to create Python programs that combine vision models for answering complex visual queries without training.","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Carl Vondrick, D\\'idac Sur\\'is, Sachit Menon","submitted_at":"2023-03-14T17:57:47Z","abstract_excerpt":"Answering visual queries is a complex task that requires both visual processing and reasoning. End-to-end models, the dominant approach for this task, do not explicitly differentiate between the two, limiting interpretability and generalization. Learning modular programs presents a promising alternative, but has proven challenging due to the difficulty of learning both the programs and modules simultaneously. We introduce ViperGPT, a framework that leverages code-generation models to compose vision-and-language models into subroutines to produce a result for any query. ViperGPT utilizes a prov"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"This simple approach requires no further training, and achieves state-of-the-art results across various complex visual tasks.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That a code-generation model can reliably produce correct, executable compositions of the provided vision modules for arbitrary queries without systematic errors in program logic or API usage.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"ViperGPT generates executable Python code to compose pre-trained vision-and-language modules into programs that answer visual queries, reaching state-of-the-art results with no additional training.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"ViperGPT uses code generation to create Python programs that combine vision models for answering complex visual queries without training.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"f15980db7861e4a53f01b42c8d6dfe8448e4e929cb3fe54574e11b2f92be62ff"},"source":{"id":"2303.08128","kind":"arxiv","version":1},"verdict":{"id":"58654800-9a63-4a21-a99b-ddaa12a324a7","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-17T18:09:01.184006Z","strongest_claim":"This simple approach requires no further training, and achieves state-of-the-art results across various complex visual tasks.","one_line_summary":"ViperGPT generates executable Python code to compose pre-trained vision-and-language modules into programs that answer visual queries, reaching state-of-the-art results with no additional training.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That a code-generation model can reliably produce correct, executable compositions of the provided vision modules for arbitrary queries without systematic errors in program logic or API usage.","pith_extraction_headline":"ViperGPT uses code generation to create Python programs that combine vision models for answering complex visual queries without training."},"references":{"count":66,"sample":[{"doi":"","year":2022,"title":"Flamingo: a visual language model for few-shot learning","work_id":"6c7c3027-f918-40e9-b9e8-c682e15c12a3","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2016,"title":"Neural module networks","work_id":"5d0a6c6e-61ca-40d3-9377-b7330e86cf81","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2019,"title":"arXiv preprint arXiv:1811.12889 , year=","work_id":"43c4bd82-57c8-4d3e-8e94-66ae1ffab8f3","ref_index":3,"cited_arxiv_id":"1811.12889","is_internal_anchor":true},{"doi":"","year":2019,"title":"The Consciousness Prior, Dec","work_id":"3d734855-ef5a-4524-afd9-bedb436a7293","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2022,"title":"Bravo, Sudhanshu Mittal, Simon Ging, and Thomas Brox","work_id":"c6fb2c6a-4031-4584-b576-d3ca0ed8686a","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":66,"snapshot_sha256":"3a390387f7b38ac80d23d44eff6c3526995982f9a376079f94e3bc33d09ec62f","internal_anchors":12},"formal_canon":{"evidence_count":1,"snapshot_sha256":"8e12ce63628226e7721f95aaf10ac650aecb66d8c42e4a5cea1dd417ee4dd854"},"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":"2303.08128","created_at":"2026-05-17T23:38:13.427272+00:00"},{"alias_kind":"arxiv_version","alias_value":"2303.08128v1","created_at":"2026-05-17T23:38:13.427272+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2303.08128","created_at":"2026-05-17T23:38:13.427272+00:00"},{"alias_kind":"pith_short_12","alias_value":"EJIVFRXU6P2M","created_at":"2026-05-18T12:33:33.725879+00:00"},{"alias_kind":"pith_short_16","alias_value":"EJIVFRXU6P2MSHA7","created_at":"2026-05-18T12:33:33.725879+00:00"},{"alias_kind":"pith_short_8","alias_value":"EJIVFRXU","created_at":"2026-05-18T12:33:33.725879+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":23,"internal_anchor_count":23,"sample":[{"citing_arxiv_id":"2404.01063","citing_title":"Chat Modeling: Interaction-Enhanced Agent Framework for Visualizing Literature-Grounded Biological Structures","ref_index":44,"is_internal_anchor":true},{"citing_arxiv_id":"2407.08101","citing_title":"What to Say and When to Say it: Live Fitness Coaching as a Testbed for Situated Interaction","ref_index":56,"is_internal_anchor":true},{"citing_arxiv_id":"2505.23678","citing_title":"Grounded Reinforcement Learning for Visual Reasoning","ref_index":56,"is_internal_anchor":true},{"citing_arxiv_id":"2512.01843","citing_title":"PhyDetEx: Detecting and Explaining the Physical Plausibility of T2V Models","ref_index":39,"is_internal_anchor":true},{"citing_arxiv_id":"2605.18697","citing_title":"PopPy: Opportunistically Exploiting Parallelism in Python Compound AI Applications","ref_index":69,"is_internal_anchor":true},{"citing_arxiv_id":"2307.06435","citing_title":"A Comprehensive Overview of Large Language Models","ref_index":226,"is_internal_anchor":true},{"citing_arxiv_id":"2605.16932","citing_title":"MORN: Metacognitive Object-Goal Regulation for Resource-Rational Long-Horizon Navigation","ref_index":44,"is_internal_anchor":true},{"citing_arxiv_id":"2512.10362","citing_title":"Visual Funnel: Resolving Contextual Blindness in Multimodal Large Language Models","ref_index":22,"is_internal_anchor":true},{"citing_arxiv_id":"2309.17421","citing_title":"The Dawn of LMMs: Preliminary Explorations with GPT-4V(ision)","ref_index":121,"is_internal_anchor":true},{"citing_arxiv_id":"2604.13046","citing_title":"A Domain-Specific Language for LLM-Driven Trigger Generation in Multimodal Data Collection","ref_index":5,"is_internal_anchor":true},{"citing_arxiv_id":"2304.15010","citing_title":"LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model","ref_index":61,"is_internal_anchor":true},{"citing_arxiv_id":"2308.11432","citing_title":"A Survey on Large Language Model based Autonomous Agents","ref_index":75,"is_internal_anchor":true},{"citing_arxiv_id":"2303.16199","citing_title":"LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init Attention","ref_index":73,"is_internal_anchor":true},{"citing_arxiv_id":"2311.10122","citing_title":"Video-LLaVA: Learning United Visual Representation by Alignment Before Projection","ref_index":76,"is_internal_anchor":true},{"citing_arxiv_id":"2303.11381","citing_title":"MM-REACT: Prompting ChatGPT for Multimodal Reasoning and Action","ref_index":25,"is_internal_anchor":true},{"citing_arxiv_id":"2312.14238","citing_title":"InternVL: Scaling up Vision Foundation Models and Aligning for Generic Visual-Linguistic Tasks","ref_index":134,"is_internal_anchor":true},{"citing_arxiv_id":"2307.05973","citing_title":"VoxPoser: Composable 3D Value Maps for Robotic Manipulation with Language Models","ref_index":137,"is_internal_anchor":true},{"citing_arxiv_id":"2311.12983","citing_title":"GAIA: a benchmark for General AI Assistants","ref_index":65,"is_internal_anchor":true},{"citing_arxiv_id":"2604.23788","citing_title":"MIRAGE: A Micro-Interaction Relational Architecture for Grounded Exploration in Multi-Figure Artworks","ref_index":31,"is_internal_anchor":true},{"citing_arxiv_id":"2604.19633","citing_title":"Time Series Augmented Generation for Financial Applications","ref_index":22,"is_internal_anchor":true},{"citing_arxiv_id":"2304.08485","citing_title":"Visual Instruction Tuning","ref_index":46,"is_internal_anchor":true},{"citing_arxiv_id":"2304.10592","citing_title":"MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large Language Models","ref_index":17,"is_internal_anchor":true},{"citing_arxiv_id":"2605.04304","citing_title":"Hierarchical Visual Agent: Managing Contexts in Joint Image-Text Space for Advanced Chart Reasoning","ref_index":51,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":1,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/EJIVFRXU6P2MSHA7YTIQRPXXRC","json":"https://pith.science/pith/EJIVFRXU6P2MSHA7YTIQRPXXRC.json","graph_json":"https://pith.science/api/pith-number/EJIVFRXU6P2MSHA7YTIQRPXXRC/graph.json","events_json":"https://pith.science/api/pith-number/EJIVFRXU6P2MSHA7YTIQRPXXRC/events.json","paper":"https://pith.science/paper/EJIVFRXU"},"agent_actions":{"view_html":"https://pith.science/pith/EJIVFRXU6P2MSHA7YTIQRPXXRC","download_json":"https://pith.science/pith/EJIVFRXU6P2MSHA7YTIQRPXXRC.json","view_paper":"https://pith.science/paper/EJIVFRXU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2303.08128&json=true","fetch_graph":"https://pith.science/api/pith-number/EJIVFRXU6P2MSHA7YTIQRPXXRC/graph.json","fetch_events":"https://pith.science/api/pith-number/EJIVFRXU6P2MSHA7YTIQRPXXRC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/EJIVFRXU6P2MSHA7YTIQRPXXRC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/EJIVFRXU6P2MSHA7YTIQRPXXRC/action/storage_attestation","attest_author":"https://pith.science/pith/EJIVFRXU6P2MSHA7YTIQRPXXRC/action/author_attestation","sign_citation":"https://pith.science/pith/EJIVFRXU6P2MSHA7YTIQRPXXRC/action/citation_signature","submit_replication":"https://pith.science/pith/EJIVFRXU6P2MSHA7YTIQRPXXRC/action/replication_record"}},"created_at":"2026-05-17T23:38:13.427272+00:00","updated_at":"2026-05-17T23:38:13.427272+00:00"}