{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:ITLQ6KESUM6MMHYBY343TENA6S","short_pith_number":"pith:ITLQ6KES","schema_version":"1.0","canonical_sha256":"44d70f2892a33cc61f01c6f9b991a0f4b86c5183bfbceddc215d9e64dae9ba9b","source":{"kind":"arxiv","id":"2503.09595","version":1},"attestation_state":"computed","paper":{"title":"PISA Experiments: Exploring Physics Post-Training for Video Diffusion Models by Watching Stuff Drop","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chenyu Li, Mike Roberts, Oscar Michel, Sainan Liu, Saining Xie, Xichen Pan","submitted_at":"2025-03-12T17:58:14Z","abstract_excerpt":"Large-scale pre-trained video generation models excel in content creation but are not reliable as physically accurate world simulators out of the box. This work studies the process of post-training these models for accurate world modeling through the lens of the simple, yet fundamental, physics task of modeling object freefall. We show state-of-the-art video generation models struggle with this basic task, despite their visually impressive outputs. To remedy this problem, we find that fine-tuning on a relatively small amount of simulated videos is effective in inducing the dropping behavior in"},"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":"2503.09595","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-03-12T17:58:14Z","cross_cats_sorted":[],"title_canon_sha256":"4f42f243c96bbe0120ca3ad437ff33021725290fcc89f1692f31d36266a8ca57","abstract_canon_sha256":"32e902a4212105e548f9cde198c1829509fb181dc564a3112d02e9c8cdbd2ae2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:29:58.867613Z","signature_b64":"lekywUDvIdulqwO5fd4GbiDOngN6iDkYahfGxsQ2h63mq5sv241Whe1roSdCyv6BlSxo0xV9CgIXVQaTScBwDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"44d70f2892a33cc61f01c6f9b991a0f4b86c5183bfbceddc215d9e64dae9ba9b","last_reissued_at":"2026-07-05T10:29:58.866825Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:29:58.866825Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"PISA Experiments: Exploring Physics Post-Training for Video Diffusion Models by Watching Stuff Drop","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chenyu Li, Mike Roberts, Oscar Michel, Sainan Liu, Saining Xie, Xichen Pan","submitted_at":"2025-03-12T17:58:14Z","abstract_excerpt":"Large-scale pre-trained video generation models excel in content creation but are not reliable as physically accurate world simulators out of the box. This work studies the process of post-training these models for accurate world modeling through the lens of the simple, yet fundamental, physics task of modeling object freefall. We show state-of-the-art video generation models struggle with this basic task, despite their visually impressive outputs. To remedy this problem, we find that fine-tuning on a relatively small amount of simulated videos is effective in inducing the dropping behavior in"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.09595","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/2503.09595/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":"2503.09595","created_at":"2026-07-05T10:29:58.866919+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.09595v1","created_at":"2026-07-05T10:29:58.866919+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.09595","created_at":"2026-07-05T10:29:58.866919+00:00"},{"alias_kind":"pith_short_12","alias_value":"ITLQ6KESUM6M","created_at":"2026-07-05T10:29:58.866919+00:00"},{"alias_kind":"pith_short_16","alias_value":"ITLQ6KESUM6MMHYB","created_at":"2026-07-05T10:29:58.866919+00:00"},{"alias_kind":"pith_short_8","alias_value":"ITLQ6KES","created_at":"2026-07-05T10:29:58.866919+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":10,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.25306","citing_title":"Physics Question Scene Graph: Fine-grained Evaluation of Physical Plausibility in Text-to-Video Generation","ref_index":28,"is_internal_anchor":false},{"citing_arxiv_id":"2606.26916","citing_title":"PhysRAG: Enhancing Physics-Awareness in Video Generation via Retrieval-Augmented Generation","ref_index":38,"is_internal_anchor":false},{"citing_arxiv_id":"2606.26694","citing_title":"PhysEditWorld: A Large-Scale Dataset Toward Physics-Editable World Models","ref_index":63,"is_internal_anchor":false},{"citing_arxiv_id":"2606.26694","citing_title":"PhysEditWorld: A Large-Scale Dataset Toward Physics-Editable World Models","ref_index":64,"is_internal_anchor":false},{"citing_arxiv_id":"2605.28816","citing_title":"Gamma-World: Generative Multi-Agent World Modeling Beyond Two Players","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2605.30346","citing_title":"YoCausal: How Far is Video Generation from World Model? A Causality Perspective","ref_index":67,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08567","citing_title":"ACWM-Phys: Investigating Generalized Physical Interaction in Action-Conditioned Video World Models","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2509.24702","citing_title":"Enhancing Physical Plausibility in Video Generation by Reasoning the Implausibility","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2510.26782","citing_title":"Cloning Deterministic Worlds: The Critical Role of Latent Geometry in Long-Horizon World Models","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08567","citing_title":"ACWM-Phys: Investigating Generalized Physical Interaction in Action-Conditioned Video World Models","ref_index":18,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ITLQ6KESUM6MMHYBY343TENA6S","json":"https://pith.science/pith/ITLQ6KESUM6MMHYBY343TENA6S.json","graph_json":"https://pith.science/api/pith-number/ITLQ6KESUM6MMHYBY343TENA6S/graph.json","events_json":"https://pith.science/api/pith-number/ITLQ6KESUM6MMHYBY343TENA6S/events.json","paper":"https://pith.science/paper/ITLQ6KES"},"agent_actions":{"view_html":"https://pith.science/pith/ITLQ6KESUM6MMHYBY343TENA6S","download_json":"https://pith.science/pith/ITLQ6KESUM6MMHYBY343TENA6S.json","view_paper":"https://pith.science/paper/ITLQ6KES","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.09595&json=true","fetch_graph":"https://pith.science/api/pith-number/ITLQ6KESUM6MMHYBY343TENA6S/graph.json","fetch_events":"https://pith.science/api/pith-number/ITLQ6KESUM6MMHYBY343TENA6S/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ITLQ6KESUM6MMHYBY343TENA6S/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ITLQ6KESUM6MMHYBY343TENA6S/action/storage_attestation","attest_author":"https://pith.science/pith/ITLQ6KESUM6MMHYBY343TENA6S/action/author_attestation","sign_citation":"https://pith.science/pith/ITLQ6KESUM6MMHYBY343TENA6S/action/citation_signature","submit_replication":"https://pith.science/pith/ITLQ6KESUM6MMHYBY343TENA6S/action/replication_record"}},"created_at":"2026-07-05T10:29:58.866919+00:00","updated_at":"2026-07-05T10:29:58.866919+00:00"}