{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:BUFKONVE5ZSCH6TJVCD6YRS76S","short_pith_number":"pith:BUFKONVE","schema_version":"1.0","canonical_sha256":"0d0aa736a4ee6423fa69a887ec465ff4a03aced626235d6f4d5e0e0badbba346","source":{"kind":"arxiv","id":"2606.07602","version":1},"attestation_state":"computed","paper":{"title":"Sample-Efficient Post-Training for LEGO Spatial-Physics Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Ge Lin Kan, Minghao Liu, Weiyang Liu, Yuhuan Yuan, Zhouliang Yu","submitted_at":"2026-05-29T09:31:25Z","abstract_excerpt":"LLM-based LEGO assembly generation requires both semantic grounding and physical feasibility. We identify a data-induced failure mode, PhysHack, in which the assemblies satisfy physical-validity constraints while producing structures that are geometrically misaligned, semantically inconsistent, or poorly calibrated. To address this challenge, we propose a model-based data selection approach that uses only a small fraction of the training data while improving physically grounded LEGO assembly generation. Building on the selected trajectories, we introduce PVPO, a sample-efficient reinforcement "},"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":"2606.07602","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-29T09:31:25Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"0d6442651eac6758e44eca91830b7ff53c69b2923e57b3807c7087f84c348bd3","abstract_canon_sha256":"33b6ba707f5f31a39dcf5d770bb5fbffc9c621e11d2b3af06c5af03db151fbf1"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-09T00:04:44.551937Z","signature_b64":"ZoxHcyk/SfxcjGpZn0ocgmUBEdYE4kDxR2+CPsVQIIRq2wSpNSL35eQuJ1YH2REkRkr170qeALJc6aVsjnm6AQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0d0aa736a4ee6423fa69a887ec465ff4a03aced626235d6f4d5e0e0badbba346","last_reissued_at":"2026-06-09T00:04:44.551420Z","signature_status":"signed_v1","first_computed_at":"2026-06-09T00:04:44.551420Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Sample-Efficient Post-Training for LEGO Spatial-Physics Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Ge Lin Kan, Minghao Liu, Weiyang Liu, Yuhuan Yuan, Zhouliang Yu","submitted_at":"2026-05-29T09:31:25Z","abstract_excerpt":"LLM-based LEGO assembly generation requires both semantic grounding and physical feasibility. We identify a data-induced failure mode, PhysHack, in which the assemblies satisfy physical-validity constraints while producing structures that are geometrically misaligned, semantically inconsistent, or poorly calibrated. To address this challenge, we propose a model-based data selection approach that uses only a small fraction of the training data while improving physically grounded LEGO assembly generation. Building on the selected trajectories, we introduce PVPO, a sample-efficient reinforcement "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.07602","kind":"arxiv","version":1},"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/2606.07602/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":"2606.07602","created_at":"2026-06-09T00:04:44.551507+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.07602v1","created_at":"2026-06-09T00:04:44.551507+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.07602","created_at":"2026-06-09T00:04:44.551507+00:00"},{"alias_kind":"pith_short_12","alias_value":"BUFKONVE5ZSC","created_at":"2026-06-09T00:04:44.551507+00:00"},{"alias_kind":"pith_short_16","alias_value":"BUFKONVE5ZSCH6TJ","created_at":"2026-06-09T00:04:44.551507+00:00"},{"alias_kind":"pith_short_8","alias_value":"BUFKONVE","created_at":"2026-06-09T00:04:44.551507+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/BUFKONVE5ZSCH6TJVCD6YRS76S","json":"https://pith.science/pith/BUFKONVE5ZSCH6TJVCD6YRS76S.json","graph_json":"https://pith.science/api/pith-number/BUFKONVE5ZSCH6TJVCD6YRS76S/graph.json","events_json":"https://pith.science/api/pith-number/BUFKONVE5ZSCH6TJVCD6YRS76S/events.json","paper":"https://pith.science/paper/BUFKONVE"},"agent_actions":{"view_html":"https://pith.science/pith/BUFKONVE5ZSCH6TJVCD6YRS76S","download_json":"https://pith.science/pith/BUFKONVE5ZSCH6TJVCD6YRS76S.json","view_paper":"https://pith.science/paper/BUFKONVE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.07602&json=true","fetch_graph":"https://pith.science/api/pith-number/BUFKONVE5ZSCH6TJVCD6YRS76S/graph.json","fetch_events":"https://pith.science/api/pith-number/BUFKONVE5ZSCH6TJVCD6YRS76S/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BUFKONVE5ZSCH6TJVCD6YRS76S/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BUFKONVE5ZSCH6TJVCD6YRS76S/action/storage_attestation","attest_author":"https://pith.science/pith/BUFKONVE5ZSCH6TJVCD6YRS76S/action/author_attestation","sign_citation":"https://pith.science/pith/BUFKONVE5ZSCH6TJVCD6YRS76S/action/citation_signature","submit_replication":"https://pith.science/pith/BUFKONVE5ZSCH6TJVCD6YRS76S/action/replication_record"}},"created_at":"2026-06-09T00:04:44.551507+00:00","updated_at":"2026-06-09T00:04:44.551507+00:00"}