{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:J2HWQRS5VUOLVOKDB73BQ44YRZ","short_pith_number":"pith:J2HWQRS5","schema_version":"1.0","canonical_sha256":"4e8f68465dad1cbab9430ff61873988e5ee0df1af61c0a95b450341870c86376","source":{"kind":"arxiv","id":"2407.07875","version":2},"attestation_state":"computed","paper":{"title":"Generative Image as Action Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.CV","cs.LG"],"primary_cat":"cs.RO","authors_text":"Mohit Shridhar, Stephen James, Yat Long Lo","submitted_at":"2024-07-10T17:41:10Z","abstract_excerpt":"Image-generation diffusion models have been fine-tuned to unlock new capabilities such as image-editing and novel view synthesis. Can we similarly unlock image-generation models for visuomotor control? We present GENIMA, a behavior-cloning agent that fine-tunes Stable Diffusion to 'draw joint-actions' as targets on RGB images. These images are fed into a controller that maps the visual targets into a sequence of joint-positions. We study GENIMA on 25 RLBench and 9 real-world manipulation tasks. We find that, by lifting actions into image-space, internet pre-trained diffusion models can generat"},"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":"2407.07875","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.RO","submitted_at":"2024-07-10T17:41:10Z","cross_cats_sorted":["cs.AI","cs.CL","cs.CV","cs.LG"],"title_canon_sha256":"25e9aeb70855171c44436afc4b7c67e66372f9a743206120aa00c4074af6c1fd","abstract_canon_sha256":"b3d0088ae817d6efc5a8ca5b11b0dd504f9b8e97e03d35fbafefd51d8da4654a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:17:45.738644Z","signature_b64":"WBQYhgQVMaZFAnKSFZj8+iXxgty5Dcx4t9duH93WY/Q4K+vYShfBFht70lCjmxQKKLuZB2X5/cxQA2s9sItYCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4e8f68465dad1cbab9430ff61873988e5ee0df1af61c0a95b450341870c86376","last_reissued_at":"2026-07-05T09:17:45.738174Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:17:45.738174Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Generative Image as Action Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.CV","cs.LG"],"primary_cat":"cs.RO","authors_text":"Mohit Shridhar, Stephen James, Yat Long Lo","submitted_at":"2024-07-10T17:41:10Z","abstract_excerpt":"Image-generation diffusion models have been fine-tuned to unlock new capabilities such as image-editing and novel view synthesis. Can we similarly unlock image-generation models for visuomotor control? We present GENIMA, a behavior-cloning agent that fine-tunes Stable Diffusion to 'draw joint-actions' as targets on RGB images. These images are fed into a controller that maps the visual targets into a sequence of joint-positions. We study GENIMA on 25 RLBench and 9 real-world manipulation tasks. We find that, by lifting actions into image-space, internet pre-trained diffusion models can generat"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.07875","kind":"arxiv","version":2},"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/2407.07875/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":"2407.07875","created_at":"2026-07-05T09:17:45.738231+00:00"},{"alias_kind":"arxiv_version","alias_value":"2407.07875v2","created_at":"2026-07-05T09:17:45.738231+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.07875","created_at":"2026-07-05T09:17:45.738231+00:00"},{"alias_kind":"pith_short_12","alias_value":"J2HWQRS5VUOL","created_at":"2026-07-05T09:17:45.738231+00:00"},{"alias_kind":"pith_short_16","alias_value":"J2HWQRS5VUOLVOKD","created_at":"2026-07-05T09:17:45.738231+00:00"},{"alias_kind":"pith_short_8","alias_value":"J2HWQRS5","created_at":"2026-07-05T09:17:45.738231+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.27079","citing_title":"ForesightSafety-VLA: A Unified Diagnostic Safety Benchmark for Vision-Language-Action Models","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2606.27079","citing_title":"ForesightSafety-VLA: A Unified Diagnostic Safety Benchmark for Vision-Language-Action Models","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2506.14135","citing_title":"GAF: Gaussian Action Field as a 4D Representation for Dynamic World Modeling in Robotic Manipulation","ref_index":53,"is_internal_anchor":false},{"citing_arxiv_id":"2503.22020","citing_title":"CoT-VLA: Visual Chain-of-Thought Reasoning for Vision-Language-Action Models","ref_index":56,"is_internal_anchor":false},{"citing_arxiv_id":"2604.04974","citing_title":"From Video to Control: A Survey of Learning Manipulation Interfaces from Temporal Visual Data","ref_index":85,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06192","citing_title":"EA-WM: Event-Aware Generative World Model with Structured Kinematic-to-Visual Action Fields","ref_index":20,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/J2HWQRS5VUOLVOKDB73BQ44YRZ","json":"https://pith.science/pith/J2HWQRS5VUOLVOKDB73BQ44YRZ.json","graph_json":"https://pith.science/api/pith-number/J2HWQRS5VUOLVOKDB73BQ44YRZ/graph.json","events_json":"https://pith.science/api/pith-number/J2HWQRS5VUOLVOKDB73BQ44YRZ/events.json","paper":"https://pith.science/paper/J2HWQRS5"},"agent_actions":{"view_html":"https://pith.science/pith/J2HWQRS5VUOLVOKDB73BQ44YRZ","download_json":"https://pith.science/pith/J2HWQRS5VUOLVOKDB73BQ44YRZ.json","view_paper":"https://pith.science/paper/J2HWQRS5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2407.07875&json=true","fetch_graph":"https://pith.science/api/pith-number/J2HWQRS5VUOLVOKDB73BQ44YRZ/graph.json","fetch_events":"https://pith.science/api/pith-number/J2HWQRS5VUOLVOKDB73BQ44YRZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/J2HWQRS5VUOLVOKDB73BQ44YRZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/J2HWQRS5VUOLVOKDB73BQ44YRZ/action/storage_attestation","attest_author":"https://pith.science/pith/J2HWQRS5VUOLVOKDB73BQ44YRZ/action/author_attestation","sign_citation":"https://pith.science/pith/J2HWQRS5VUOLVOKDB73BQ44YRZ/action/citation_signature","submit_replication":"https://pith.science/pith/J2HWQRS5VUOLVOKDB73BQ44YRZ/action/replication_record"}},"created_at":"2026-07-05T09:17:45.738231+00:00","updated_at":"2026-07-05T09:17:45.738231+00:00"}