{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:GGYGFLDRBY3ACVP3KIOYFIRHNL","short_pith_number":"pith:GGYGFLDR","canonical_record":{"source":{"id":"2512.10891","kind":"arxiv","version":5},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.RO","submitted_at":"2025-12-11T18:20:49Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"86d3da3140c7bd9842e6cf20423d3b28835c7f337fb72857a7c57871c93fabea","abstract_canon_sha256":"e738704ef46298c2c8c1dcc996f4f33d649ea22e41941c7181f81aef4505bf7a"},"schema_version":"1.0"},"canonical_sha256":"31b062ac710e360155fb521d82a2276afa39201130188326c6dfbc4ad1c0f427","source":{"kind":"arxiv","id":"2512.10891","version":5},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2512.10891","created_at":"2026-05-20T01:05:03Z"},{"alias_kind":"arxiv_version","alias_value":"2512.10891v5","created_at":"2026-05-20T01:05:03Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2512.10891","created_at":"2026-05-20T01:05:03Z"},{"alias_kind":"pith_short_12","alias_value":"GGYGFLDRBY3A","created_at":"2026-05-20T01:05:03Z"},{"alias_kind":"pith_short_16","alias_value":"GGYGFLDRBY3ACVP3","created_at":"2026-05-20T01:05:03Z"},{"alias_kind":"pith_short_8","alias_value":"GGYGFLDR","created_at":"2026-05-20T01:05:03Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:GGYGFLDRBY3ACVP3KIOYFIRHNL","target":"record","payload":{"canonical_record":{"source":{"id":"2512.10891","kind":"arxiv","version":5},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.RO","submitted_at":"2025-12-11T18:20:49Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"86d3da3140c7bd9842e6cf20423d3b28835c7f337fb72857a7c57871c93fabea","abstract_canon_sha256":"e738704ef46298c2c8c1dcc996f4f33d649ea22e41941c7181f81aef4505bf7a"},"schema_version":"1.0"},"canonical_sha256":"31b062ac710e360155fb521d82a2276afa39201130188326c6dfbc4ad1c0f427","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T01:05:03.798483Z","signature_b64":"hUrRqfTxb83eLAYKam8OSHoO/bpkZpn/TlVWKQ8g5czzO+AldDJKyN69lFa9qgVBoNMCv5KNndrNSvSv0LyrBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"31b062ac710e360155fb521d82a2276afa39201130188326c6dfbc4ad1c0f427","last_reissued_at":"2026-05-20T01:05:03.797303Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T01:05:03.797303Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2512.10891","source_version":5,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-20T01:05:03Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"3DvKzdGyruuKfwaUdMb2FjUif6YwiaEy8Om6O48YD4jqkTpbQjDkvqKsw4LUw1AvFDiq7+Lhs/DMWWJSlXT8DQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-31T05:38:53.824614Z"},"content_sha256":"2a40cde3e13cbbe9b627f0febd7030e2225fb80a9a83acdfb8f13cc652ad0190","schema_version":"1.0","event_id":"sha256:2a40cde3e13cbbe9b627f0febd7030e2225fb80a9a83acdfb8f13cc652ad0190"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:GGYGFLDRBY3ACVP3KIOYFIRHNL","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Iterative Compositional Data Generation for Robot Control","license":"http://creativecommons.org/licenses/by/4.0/","headline":"A diffusion transformer factorizes robotic transitions into semantic components and generates data for unseen task combinations after limited training.","cross_cats":["cs.LG"],"primary_cat":"cs.RO","authors_text":"Anh-Quan Pham, Dani S. Bassett, Eric Eaton, Jorge Mendez-Mendez, Marcel Hussing, Shubhankar P. Patankar","submitted_at":"2025-12-11T18:20:49Z","abstract_excerpt":"Collecting robotic manipulation data is expensive, making it impractical to acquire demonstrations for the combinatorially large space of tasks that arise in multi-object, multi-robot, and multi-environment settings. While recent generative models can synthesize useful data for individual tasks, they do not exploit the compositional structure of robotic domains and struggle to generalize to unseen task combinations. We propose a semantic compositional diffusion transformer that factorizes transitions into robot-, object-, obstacle-, and objective-specific components and learns their interactio"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Once trained on a limited subset of tasks, our model can zero-shot generate high-quality transitions from which we can learn control policies for unseen task combinations. Our approach substantially improves zero-shot performance over monolithic and hard-coded compositional baselines, ultimately solving nearly all held-out tasks and demonstrating the emergence of meaningful compositional structure in the learned representations.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The robotic domain possesses a clean compositional structure that can be factorized into robot-, object-, obstacle-, and objective-specific components whose interactions are sufficiently captured by attention for reliable zero-shot generalization to arbitrary unseen combinations.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"A compositional diffusion model generates zero-shot data for unseen robotic task combinations and iteratively improves via RL validation, solving nearly all held-out tasks.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A diffusion transformer factorizes robotic transitions into semantic components and generates data for unseen task combinations after limited training.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"7e2ee8f82ac0a836bac98d95dba58c7cd105b4b368258d5ef02f7feee1f3e6e3"},"source":{"id":"2512.10891","kind":"arxiv","version":5},"verdict":{"id":"2e3e9a62-1aef-458b-a26d-4e686418ccbf","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-16T22:56:19.881389Z","strongest_claim":"Once trained on a limited subset of tasks, our model can zero-shot generate high-quality transitions from which we can learn control policies for unseen task combinations. Our approach substantially improves zero-shot performance over monolithic and hard-coded compositional baselines, ultimately solving nearly all held-out tasks and demonstrating the emergence of meaningful compositional structure in the learned representations.","one_line_summary":"A compositional diffusion model generates zero-shot data for unseen robotic task combinations and iteratively improves via RL validation, solving nearly all held-out tasks.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The robotic domain possesses a clean compositional structure that can be factorized into robot-, object-, obstacle-, and objective-specific components whose interactions are sufficiently captured by attention for reliable zero-shot generalization to arbitrary unseen combinations.","pith_extraction_headline":"A diffusion transformer factorizes robotic transitions into semantic components and generates data for unseen task combinations after limited training."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2512.10891/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":2,"snapshot_sha256":"ba54d345a4fb2a8c4cb82059c9db07de4babbd30cb975d6058a1f98d927c31cd"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"2e3e9a62-1aef-458b-a26d-4e686418ccbf"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-20T01:05:03Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"jPk1Rht4WsCxuu4VK9RtrJkZF0iQYrys9OKN4vqm3sGNA98KssTbIA6AUnQK8elj+lUVKZVkzA4gCkHkVgOVAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-31T05:38:53.825476Z"},"content_sha256":"d0a2cd91fcf8a9d9ab02cfd55f4439fef986cfff05e970627b4cc8203075298a","schema_version":"1.0","event_id":"sha256:d0a2cd91fcf8a9d9ab02cfd55f4439fef986cfff05e970627b4cc8203075298a"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/GGYGFLDRBY3ACVP3KIOYFIRHNL/bundle.json","state_url":"https://pith.science/pith/GGYGFLDRBY3ACVP3KIOYFIRHNL/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/GGYGFLDRBY3ACVP3KIOYFIRHNL/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-05-31T05:38:53Z","links":{"resolver":"https://pith.science/pith/GGYGFLDRBY3ACVP3KIOYFIRHNL","bundle":"https://pith.science/pith/GGYGFLDRBY3ACVP3KIOYFIRHNL/bundle.json","state":"https://pith.science/pith/GGYGFLDRBY3ACVP3KIOYFIRHNL/state.json","well_known_bundle":"https://pith.science/.well-known/pith/GGYGFLDRBY3ACVP3KIOYFIRHNL/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:GGYGFLDRBY3ACVP3KIOYFIRHNL","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"e738704ef46298c2c8c1dcc996f4f33d649ea22e41941c7181f81aef4505bf7a","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.RO","submitted_at":"2025-12-11T18:20:49Z","title_canon_sha256":"86d3da3140c7bd9842e6cf20423d3b28835c7f337fb72857a7c57871c93fabea"},"schema_version":"1.0","source":{"id":"2512.10891","kind":"arxiv","version":5}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2512.10891","created_at":"2026-05-20T01:05:03Z"},{"alias_kind":"arxiv_version","alias_value":"2512.10891v5","created_at":"2026-05-20T01:05:03Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2512.10891","created_at":"2026-05-20T01:05:03Z"},{"alias_kind":"pith_short_12","alias_value":"GGYGFLDRBY3A","created_at":"2026-05-20T01:05:03Z"},{"alias_kind":"pith_short_16","alias_value":"GGYGFLDRBY3ACVP3","created_at":"2026-05-20T01:05:03Z"},{"alias_kind":"pith_short_8","alias_value":"GGYGFLDR","created_at":"2026-05-20T01:05:03Z"}],"graph_snapshots":[{"event_id":"sha256:d0a2cd91fcf8a9d9ab02cfd55f4439fef986cfff05e970627b4cc8203075298a","target":"graph","created_at":"2026-05-20T01:05:03Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"Once trained on a limited subset of tasks, our model can zero-shot generate high-quality transitions from which we can learn control policies for unseen task combinations. Our approach substantially improves zero-shot performance over monolithic and hard-coded compositional baselines, ultimately solving nearly all held-out tasks and demonstrating the emergence of meaningful compositional structure in the learned representations."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"The robotic domain possesses a clean compositional structure that can be factorized into robot-, object-, obstacle-, and objective-specific components whose interactions are sufficiently captured by attention for reliable zero-shot generalization to arbitrary unseen combinations."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"A compositional diffusion model generates zero-shot data for unseen robotic task combinations and iteratively improves via RL validation, solving nearly all held-out tasks."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"A diffusion transformer factorizes robotic transitions into semantic components and generates data for unseen task combinations after limited training."}],"snapshot_sha256":"7e2ee8f82ac0a836bac98d95dba58c7cd105b4b368258d5ef02f7feee1f3e6e3"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"ba54d345a4fb2a8c4cb82059c9db07de4babbd30cb975d6058a1f98d927c31cd"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2512.10891/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Collecting robotic manipulation data is expensive, making it impractical to acquire demonstrations for the combinatorially large space of tasks that arise in multi-object, multi-robot, and multi-environment settings. While recent generative models can synthesize useful data for individual tasks, they do not exploit the compositional structure of robotic domains and struggle to generalize to unseen task combinations. We propose a semantic compositional diffusion transformer that factorizes transitions into robot-, object-, obstacle-, and objective-specific components and learns their interactio","authors_text":"Anh-Quan Pham, Dani S. Bassett, Eric Eaton, Jorge Mendez-Mendez, Marcel Hussing, Shubhankar P. Patankar","cross_cats":["cs.LG"],"headline":"A diffusion transformer factorizes robotic transitions into semantic components and generates data for unseen task combinations after limited training.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.RO","submitted_at":"2025-12-11T18:20:49Z","title":"Iterative Compositional Data Generation for Robot Control"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2512.10891","kind":"arxiv","version":5},"verdict":{"created_at":"2026-05-16T22:56:19.881389Z","id":"2e3e9a62-1aef-458b-a26d-4e686418ccbf","model_set":{"reader":"grok-4.3"},"one_line_summary":"A compositional diffusion model generates zero-shot data for unseen robotic task combinations and iteratively improves via RL validation, solving nearly all held-out tasks.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"A diffusion transformer factorizes robotic transitions into semantic components and generates data for unseen task combinations after limited training.","strongest_claim":"Once trained on a limited subset of tasks, our model can zero-shot generate high-quality transitions from which we can learn control policies for unseen task combinations. Our approach substantially improves zero-shot performance over monolithic and hard-coded compositional baselines, ultimately solving nearly all held-out tasks and demonstrating the emergence of meaningful compositional structure in the learned representations.","weakest_assumption":"The robotic domain possesses a clean compositional structure that can be factorized into robot-, object-, obstacle-, and objective-specific components whose interactions are sufficiently captured by attention for reliable zero-shot generalization to arbitrary unseen combinations."}},"verdict_id":"2e3e9a62-1aef-458b-a26d-4e686418ccbf"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:2a40cde3e13cbbe9b627f0febd7030e2225fb80a9a83acdfb8f13cc652ad0190","target":"record","created_at":"2026-05-20T01:05:03Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"e738704ef46298c2c8c1dcc996f4f33d649ea22e41941c7181f81aef4505bf7a","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.RO","submitted_at":"2025-12-11T18:20:49Z","title_canon_sha256":"86d3da3140c7bd9842e6cf20423d3b28835c7f337fb72857a7c57871c93fabea"},"schema_version":"1.0","source":{"id":"2512.10891","kind":"arxiv","version":5}},"canonical_sha256":"31b062ac710e360155fb521d82a2276afa39201130188326c6dfbc4ad1c0f427","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"31b062ac710e360155fb521d82a2276afa39201130188326c6dfbc4ad1c0f427","first_computed_at":"2026-05-20T01:05:03.797303Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-20T01:05:03.797303Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"hUrRqfTxb83eLAYKam8OSHoO/bpkZpn/TlVWKQ8g5czzO+AldDJKyN69lFa9qgVBoNMCv5KNndrNSvSv0LyrBA==","signature_status":"signed_v1","signed_at":"2026-05-20T01:05:03.798483Z","signed_message":"canonical_sha256_bytes"},"source_id":"2512.10891","source_kind":"arxiv","source_version":5}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:2a40cde3e13cbbe9b627f0febd7030e2225fb80a9a83acdfb8f13cc652ad0190","sha256:d0a2cd91fcf8a9d9ab02cfd55f4439fef986cfff05e970627b4cc8203075298a"],"state_sha256":"6bc66db21576e79bbeb8c6825b1056caccdadce19f55cea3f088bd0e76ab4883"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"UsHO4B0Ud0Gtp1xlo+RresFurclr6ZwgReh0LhFoRRIJeFOF/QBS5t16hVkuOIbKR+ATYrdnXJLQnk+btvitDg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-31T05:38:53.830042Z","bundle_sha256":"82ec060165370e0511ac1796315c30ce29a2cd7756a7b19af3915cd49b9e077c"}}