{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:3ICG2E6RA6GZXUIXEZKABTNWL7","short_pith_number":"pith:3ICG2E6R","canonical_record":{"source":{"id":"2605.13775","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.RO","submitted_at":"2026-05-13T16:54:36Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"328a3b6e0c3d344bb503df46cd324d610552c22aaa5b03aef663bacb9a2258cf","abstract_canon_sha256":"cfd3e2b32bbeda56f2440d71939ec6e7c4400f2a89d85d9495f953b2210152f8"},"schema_version":"1.0"},"canonical_sha256":"da046d13d1078d9bd117265400cdb65fc4100408c09e0b4979a7aab90b2a236a","source":{"kind":"arxiv","id":"2605.13775","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.13775","created_at":"2026-05-18T02:44:15Z"},{"alias_kind":"arxiv_version","alias_value":"2605.13775v1","created_at":"2026-05-18T02:44:15Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.13775","created_at":"2026-05-18T02:44:15Z"},{"alias_kind":"pith_short_12","alias_value":"3ICG2E6RA6GZ","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"3ICG2E6RA6GZXUIX","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"3ICG2E6R","created_at":"2026-05-18T12:33:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:3ICG2E6RA6GZXUIXEZKABTNWL7","target":"record","payload":{"canonical_record":{"source":{"id":"2605.13775","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.RO","submitted_at":"2026-05-13T16:54:36Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"328a3b6e0c3d344bb503df46cd324d610552c22aaa5b03aef663bacb9a2258cf","abstract_canon_sha256":"cfd3e2b32bbeda56f2440d71939ec6e7c4400f2a89d85d9495f953b2210152f8"},"schema_version":"1.0"},"canonical_sha256":"da046d13d1078d9bd117265400cdb65fc4100408c09e0b4979a7aab90b2a236a","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T02:44:15.926931Z","signature_b64":"F/zejwfet5NqbzbB2kcCVG5NgXIoqNK60pUaxW9/AeD6MSljK8CvuS/U9zpijEI1h7kM5bSQpXuIuZ3Ihn0TBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"da046d13d1078d9bd117265400cdb65fc4100408c09e0b4979a7aab90b2a236a","last_reissued_at":"2026-05-18T02:44:15.926422Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T02:44:15.926422Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.13775","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T02:44:15Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"xlvx/B0CsbSvPT0UrlUYYZvBTay2IulDWC0JiL7IqBAAYqeNPYO/MnKGzUxPRdDu25rCTHfk0tVq+prajrGlBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-09T08:19:55.520487Z"},"content_sha256":"2f1220cd27988e117198296f4a5b95d0d62041e563834874bae2c01eda9f9cd8","schema_version":"1.0","event_id":"sha256:2f1220cd27988e117198296f4a5b95d0d62041e563834874bae2c01eda9f9cd8"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:3ICG2E6RA6GZXUIXEZKABTNWL7","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"RoboEvolve: Co-Evolving Planner-Simulator for Robotic Manipulation with Limited Data","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"A co-evolutionary VLM-VGM loop on 500 unlabeled images raises planner success by 30 points and simulator success by 48 percent while beating fully supervised baselines.","cross_cats":["cs.CV"],"primary_cat":"cs.RO","authors_text":"Harold Haodong Chen, Sirui Chen, Wenhang Ge, Ying-Cong Chen, Yingjie Xu","submitted_at":"2026-05-13T16:54:36Z","abstract_excerpt":"The scalability of robotic manipulation is fundamentally bottlenecked by the scarcity of task-aligned physical interaction data. While vision-language models (VLMs) and video generation models (VGMs) hold promise for autonomous data synthesis, they suffer from semantic-spatial misalignment and physical hallucinations, respectively. To bridge this gap, we introduce RoboEvolve, a novel framework that couples a VLM planner and a VGM simulator into a mutually reinforcing co-evolutionary loop. Operating purely on unlabeled seed images, RoboEvolve leverages a cognitive-inspired dual-phase mechanism:"},"claims":{"count":3,"items":[{"kind":"strongest_claim","text":"RoboEvolve elevates base planners by 30 absolute points and amplifies simulator success by 48% on average, surpassing fully supervised baselines with merely 500 unlabeled seeds—a 50x reduction—while demonstrating robust continual learning without catastrophic forgetting.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the semantic-controlled multi-granular reward and nighttime mining of near-miss failures can enforce physical grounding and eliminate hallucinations in the VGM without any external verification or real-robot feedback.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"A co-evolutionary VLM-VGM loop on 500 unlabeled images raises planner success by 30 points and simulator success by 48 percent while beating fully supervised baselines.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"}],"snapshot_sha256":"e3005a924468904730e5c5282061ea2e1683f7edb2555c966557470a4f64ca05"},"source":{"id":"2605.13775","kind":"arxiv","version":1},"verdict":{"id":"698432ae-044f-499b-8d9c-ad4b92d3ffda","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-14T17:51:46.495579Z","strongest_claim":"RoboEvolve elevates base planners by 30 absolute points and amplifies simulator success by 48% on average, surpassing fully supervised baselines with merely 500 unlabeled seeds—a 50x reduction—while demonstrating robust continual learning without catastrophic forgetting.","one_line_summary":"A co-evolutionary VLM-VGM loop on 500 unlabeled images raises planner success by 30 points and simulator success by 48 percent while beating fully supervised baselines.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the semantic-controlled multi-granular reward and nighttime mining of near-miss failures can enforce physical grounding and eliminate hallucinations in the VGM without any external verification or real-robot feedback.","pith_extraction_headline":""},"references":{"count":83,"sample":[{"doi":"","year":null,"title":"Mind-v: Hierarchical video generation for long-horizon robotic manipulation with rl-based physical alignment","work_id":"4a9bf4e3-cdb2-4f3d-a0d5-9a5d730eef6f","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"CogVideoX: Text-to-Video Diffusion Models with An Expert Transformer","work_id":"f38fc088-12aa-4bf4-9ecd-08d3e797ccb7","ref_index":2,"cited_arxiv_id":"2408.06072","is_internal_anchor":true},{"doi":"","year":2024,"title":"European conference on computer vision , pages=","work_id":"4bc13a3e-6c33-434d-92e8-e61c499a776f","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"HunyuanVideo: A Systematic Framework For Large Video Generative Models","work_id":"881efa7e-7e73-4c66-9cc3-2803e551061c","ref_index":4,"cited_arxiv_id":"2412.03603","is_internal_anchor":true},{"doi":"","year":null,"title":"World Action Models are Zero-shot Policies","work_id":"9a85fc69-74df-450e-94cd-69d186e9e830","ref_index":5,"cited_arxiv_id":"2602.15922","is_internal_anchor":true}],"resolved_work":83,"snapshot_sha256":"3855d36bd55a549ac6a762c270f2630dfcfe47942979559367263fc7be94b689","internal_anchors":27},"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"},"verdict_id":"698432ae-044f-499b-8d9c-ad4b92d3ffda"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-18T02:44:15Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"VcFQWA8HBkEyLNQlF445kdbR/vd61Ex9b4LIBq/KA901nYGaTkxxCVgjb8zfiAZ5FAP8/8Q/0fgmxTTX3tLVCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-09T08:19:55.521898Z"},"content_sha256":"060f8a87af1ad9a6f343cef22fdda65a23021330efc3d1b5659d4c638e188072","schema_version":"1.0","event_id":"sha256:060f8a87af1ad9a6f343cef22fdda65a23021330efc3d1b5659d4c638e188072"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/3ICG2E6RA6GZXUIXEZKABTNWL7/bundle.json","state_url":"https://pith.science/pith/3ICG2E6RA6GZXUIXEZKABTNWL7/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/3ICG2E6RA6GZXUIXEZKABTNWL7/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":"2026-06-09T08:19:55Z","links":{"resolver":"https://pith.science/pith/3ICG2E6RA6GZXUIXEZKABTNWL7","bundle":"https://pith.science/pith/3ICG2E6RA6GZXUIXEZKABTNWL7/bundle.json","state":"https://pith.science/pith/3ICG2E6RA6GZXUIXEZKABTNWL7/state.json","well_known_bundle":"https://pith.science/.well-known/pith/3ICG2E6RA6GZXUIXEZKABTNWL7/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:3ICG2E6RA6GZXUIXEZKABTNWL7","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":"cfd3e2b32bbeda56f2440d71939ec6e7c4400f2a89d85d9495f953b2210152f8","cross_cats_sorted":["cs.CV"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.RO","submitted_at":"2026-05-13T16:54:36Z","title_canon_sha256":"328a3b6e0c3d344bb503df46cd324d610552c22aaa5b03aef663bacb9a2258cf"},"schema_version":"1.0","source":{"id":"2605.13775","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.13775","created_at":"2026-05-18T02:44:15Z"},{"alias_kind":"arxiv_version","alias_value":"2605.13775v1","created_at":"2026-05-18T02:44:15Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.13775","created_at":"2026-05-18T02:44:15Z"},{"alias_kind":"pith_short_12","alias_value":"3ICG2E6RA6GZ","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"3ICG2E6RA6GZXUIX","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"3ICG2E6R","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:060f8a87af1ad9a6f343cef22fdda65a23021330efc3d1b5659d4c638e188072","target":"graph","created_at":"2026-05-18T02:44:15Z","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":3,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"RoboEvolve elevates base planners by 30 absolute points and amplifies simulator success by 48% on average, surpassing fully supervised baselines with merely 500 unlabeled seeds—a 50x reduction—while demonstrating robust continual learning without catastrophic forgetting."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That the semantic-controlled multi-granular reward and nighttime mining of near-miss failures can enforce physical grounding and eliminate hallucinations in the VGM without any external verification or real-robot feedback."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"A co-evolutionary VLM-VGM loop on 500 unlabeled images raises planner success by 30 points and simulator success by 48 percent while beating fully supervised baselines."}],"snapshot_sha256":"e3005a924468904730e5c5282061ea2e1683f7edb2555c966557470a4f64ca05"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"paper":{"abstract_excerpt":"The scalability of robotic manipulation is fundamentally bottlenecked by the scarcity of task-aligned physical interaction data. While vision-language models (VLMs) and video generation models (VGMs) hold promise for autonomous data synthesis, they suffer from semantic-spatial misalignment and physical hallucinations, respectively. To bridge this gap, we introduce RoboEvolve, a novel framework that couples a VLM planner and a VGM simulator into a mutually reinforcing co-evolutionary loop. Operating purely on unlabeled seed images, RoboEvolve leverages a cognitive-inspired dual-phase mechanism:","authors_text":"Harold Haodong Chen, Sirui Chen, Wenhang Ge, Ying-Cong Chen, Yingjie Xu","cross_cats":["cs.CV"],"headline":"A co-evolutionary VLM-VGM loop on 500 unlabeled images raises planner success by 30 points and simulator success by 48 percent while beating fully supervised baselines.","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.RO","submitted_at":"2026-05-13T16:54:36Z","title":"RoboEvolve: Co-Evolving Planner-Simulator for Robotic Manipulation with Limited Data"},"references":{"count":83,"internal_anchors":27,"resolved_work":83,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"Mind-v: Hierarchical video generation for long-horizon robotic manipulation with rl-based physical alignment","work_id":"4a9bf4e3-cdb2-4f3d-a0d5-9a5d730eef6f","year":null},{"cited_arxiv_id":"2408.06072","doi":"","is_internal_anchor":true,"ref_index":2,"title":"CogVideoX: Text-to-Video Diffusion Models with An Expert Transformer","work_id":"f38fc088-12aa-4bf4-9ecd-08d3e797ccb7","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"European conference on computer vision , pages=","work_id":"4bc13a3e-6c33-434d-92e8-e61c499a776f","year":2024},{"cited_arxiv_id":"2412.03603","doi":"","is_internal_anchor":true,"ref_index":4,"title":"HunyuanVideo: A Systematic Framework For Large Video Generative Models","work_id":"881efa7e-7e73-4c66-9cc3-2803e551061c","year":null},{"cited_arxiv_id":"2602.15922","doi":"","is_internal_anchor":true,"ref_index":5,"title":"World Action Models are Zero-shot Policies","work_id":"9a85fc69-74df-450e-94cd-69d186e9e830","year":null}],"snapshot_sha256":"3855d36bd55a549ac6a762c270f2630dfcfe47942979559367263fc7be94b689"},"source":{"id":"2605.13775","kind":"arxiv","version":1},"verdict":{"created_at":"2026-05-14T17:51:46.495579Z","id":"698432ae-044f-499b-8d9c-ad4b92d3ffda","model_set":{"reader":"grok-4.3"},"one_line_summary":"A co-evolutionary VLM-VGM loop on 500 unlabeled images raises planner success by 30 points and simulator success by 48 percent while beating fully supervised baselines.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"","strongest_claim":"RoboEvolve elevates base planners by 30 absolute points and amplifies simulator success by 48% on average, surpassing fully supervised baselines with merely 500 unlabeled seeds—a 50x reduction—while demonstrating robust continual learning without catastrophic forgetting.","weakest_assumption":"That the semantic-controlled multi-granular reward and nighttime mining of near-miss failures can enforce physical grounding and eliminate hallucinations in the VGM without any external verification or real-robot feedback."}},"verdict_id":"698432ae-044f-499b-8d9c-ad4b92d3ffda"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:2f1220cd27988e117198296f4a5b95d0d62041e563834874bae2c01eda9f9cd8","target":"record","created_at":"2026-05-18T02:44:15Z","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":"cfd3e2b32bbeda56f2440d71939ec6e7c4400f2a89d85d9495f953b2210152f8","cross_cats_sorted":["cs.CV"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.RO","submitted_at":"2026-05-13T16:54:36Z","title_canon_sha256":"328a3b6e0c3d344bb503df46cd324d610552c22aaa5b03aef663bacb9a2258cf"},"schema_version":"1.0","source":{"id":"2605.13775","kind":"arxiv","version":1}},"canonical_sha256":"da046d13d1078d9bd117265400cdb65fc4100408c09e0b4979a7aab90b2a236a","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"da046d13d1078d9bd117265400cdb65fc4100408c09e0b4979a7aab90b2a236a","first_computed_at":"2026-05-18T02:44:15.926422Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T02:44:15.926422Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"F/zejwfet5NqbzbB2kcCVG5NgXIoqNK60pUaxW9/AeD6MSljK8CvuS/U9zpijEI1h7kM5bSQpXuIuZ3Ihn0TBA==","signature_status":"signed_v1","signed_at":"2026-05-18T02:44:15.926931Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.13775","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:2f1220cd27988e117198296f4a5b95d0d62041e563834874bae2c01eda9f9cd8","sha256:060f8a87af1ad9a6f343cef22fdda65a23021330efc3d1b5659d4c638e188072"],"state_sha256":"c1d7da523b79535f68d6e5d87a1b3cd792b10977998a9f11b59d49834f3ecbbf"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"oDuRUMEk3Gs11o2XonIL7QdTs/2pnfHLyEz3sC++//XFPY1R40oezw5WrLRLpQr2St3ooDRsOo8xMK68Bp8jAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-09T08:19:55.528108Z","bundle_sha256":"e7823d26919340ac232d8eca576f88b11636dbd178e833b39c82bba2a676e862"}}