{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:42LIIEDICQJM5JBVO3PXB7OH6L","short_pith_number":"pith:42LIIEDI","schema_version":"1.0","canonical_sha256":"e6968410681412cea43576df70fdc7f2da4ae67326711be3fd90004d96260e09","source":{"kind":"arxiv","id":"2511.15684","version":2},"attestation_state":"computed","paper":{"title":"Walrus: A Cross-Domain Foundation Model for Continuum Dynamics","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CE"],"primary_cat":"cs.LG","authors_text":"Alberto Bietti, Bruno Regaldo-Saint Blancard, Cristiana Diaconu, Francois Lanusse, Francois Rozet, Geraud Krawezik, Hadi Sotoudeh, Irina Espejo, Jeff Shen, Kaze W. K. Wong, Keiya Hirashima, Kyunghyun Cho, Liam Parker, Lucas Meyer, Mariel Pettee, Michael McCabe, Miles Cranmer, Payel Mukhopadhyay, Rio Fear, Ruben Ohana, Rudy Morel, Shirley Ho, Siavash Golkar, Tanya Marwah, Tom Hehir","submitted_at":"2025-11-19T18:36:03Z","abstract_excerpt":"Foundation models have transformed machine learning for language and vision, but achieving comparable impact in physical simulation remains a challenge. Data heterogeneity and unstable long-term dynamics inhibit learning from sufficiently diverse dynamics, while varying resolutions and dimensionalities challenge efficient training on modern hardware. Through empirical and theoretical analysis, we incorporate new approaches to mitigate these obstacles, including a harmonic-analysis-based stabilization method, load-balanced distributed 2D and 3D training strategies, and compute-adaptive tokeniza"},"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":"2511.15684","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-11-19T18:36:03Z","cross_cats_sorted":["cs.AI","cs.CE"],"title_canon_sha256":"90b3448c2efb05d599b09643881c77ed4c0fc3440d7fceee86573f5ea1369a56","abstract_canon_sha256":"712023d44869b09c0fac271fcb6704c043b91f0187e1e58a519c9bf799d32c8b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-07T02:18:30.462036Z","signature_b64":"pOaderotnGITnSv8fqBz1pzYZCPGklz5MJaGq9kQRKWpJkrzyIPoXsV+suLxmJwAgkPvm3OteXTws9zSBA2eAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e6968410681412cea43576df70fdc7f2da4ae67326711be3fd90004d96260e09","last_reissued_at":"2026-07-07T02:18:30.461118Z","signature_status":"signed_v1","first_computed_at":"2026-07-07T02:18:30.461118Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Walrus: A Cross-Domain Foundation Model for Continuum Dynamics","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CE"],"primary_cat":"cs.LG","authors_text":"Alberto Bietti, Bruno Regaldo-Saint Blancard, Cristiana Diaconu, Francois Lanusse, Francois Rozet, Geraud Krawezik, Hadi Sotoudeh, Irina Espejo, Jeff Shen, Kaze W. K. Wong, Keiya Hirashima, Kyunghyun Cho, Liam Parker, Lucas Meyer, Mariel Pettee, Michael McCabe, Miles Cranmer, Payel Mukhopadhyay, Rio Fear, Ruben Ohana, Rudy Morel, Shirley Ho, Siavash Golkar, Tanya Marwah, Tom Hehir","submitted_at":"2025-11-19T18:36:03Z","abstract_excerpt":"Foundation models have transformed machine learning for language and vision, but achieving comparable impact in physical simulation remains a challenge. Data heterogeneity and unstable long-term dynamics inhibit learning from sufficiently diverse dynamics, while varying resolutions and dimensionalities challenge efficient training on modern hardware. Through empirical and theoretical analysis, we incorporate new approaches to mitigate these obstacles, including a harmonic-analysis-based stabilization method, load-balanced distributed 2D and 3D training strategies, and compute-adaptive tokeniza"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2511.15684","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/2511.15684/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":"2511.15684","created_at":"2026-07-07T02:18:30.461247+00:00"},{"alias_kind":"arxiv_version","alias_value":"2511.15684v2","created_at":"2026-07-07T02:18:30.461247+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2511.15684","created_at":"2026-07-07T02:18:30.461247+00:00"},{"alias_kind":"pith_short_12","alias_value":"42LIIEDICQJM","created_at":"2026-07-07T02:18:30.461247+00:00"},{"alias_kind":"pith_short_16","alias_value":"42LIIEDICQJM5JBV","created_at":"2026-07-07T02:18:30.461247+00:00"},{"alias_kind":"pith_short_8","alias_value":"42LIIEDI","created_at":"2026-07-07T02:18:30.461247+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":13,"internal_anchor_count":13,"sample":[{"citing_arxiv_id":"2606.26421","citing_title":"Otter Weather: Skillful and Computationally Efficient Medium-Range Weather Forecasting","ref_index":24,"is_internal_anchor":true},{"citing_arxiv_id":"2606.15015","citing_title":"NEXUS: Neural Energy Fields for Physically Consistent Contact-Rich 3D Object Dynamics","ref_index":9,"is_internal_anchor":true},{"citing_arxiv_id":"2605.28851","citing_title":"Towards a Foundation Model for the Martian Atmosphere","ref_index":158,"is_internal_anchor":true},{"citing_arxiv_id":"2605.25949","citing_title":"Small Models, Strong Priors: Architectural Inductive Bias for Parameter-Efficient Neural PDE Solvers","ref_index":34,"is_internal_anchor":true},{"citing_arxiv_id":"2606.11657","citing_title":"Sparse probes and murky physics: a case study of interpretability challenges in a foundation model for continuum dynamics","ref_index":2,"is_internal_anchor":true},{"citing_arxiv_id":"2605.15793","citing_title":"AOT-POT: Adaptive Operator Transformation for Large-Scale PDE Pre-training","ref_index":40,"is_internal_anchor":true},{"citing_arxiv_id":"2605.17968","citing_title":"Function graph transformers universally approximate operators between function spaces","ref_index":79,"is_internal_anchor":true},{"citing_arxiv_id":"2605.15399","citing_title":"Breakeven complexity: A new perspective on neural partial differential equation solvers","ref_index":21,"is_internal_anchor":true},{"citing_arxiv_id":"2601.10791","citing_title":"OmniMol: Transferring Particle Physics Knowledge to Molecular Dynamics with Point-Edge Transformers","ref_index":27,"is_internal_anchor":true},{"citing_arxiv_id":"2602.18679","citing_title":"Transformers for dynamical systems learn transfer operators in-context","ref_index":22,"is_internal_anchor":true},{"citing_arxiv_id":"2605.14546","citing_title":"Discovering Physical Directions in Weight Space: Composing Neural PDE Experts","ref_index":19,"is_internal_anchor":true},{"citing_arxiv_id":"2605.10159","citing_title":"jNO: A JAX Library for Neural Operator and Foundation Model Training","ref_index":8,"is_internal_anchor":true},{"citing_arxiv_id":"2605.04832","citing_title":"Replay-Based Continual Learning for Physics-Informed Neural Operators","ref_index":16,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/42LIIEDICQJM5JBVO3PXB7OH6L","json":"https://pith.science/pith/42LIIEDICQJM5JBVO3PXB7OH6L.json","graph_json":"https://pith.science/api/pith-number/42LIIEDICQJM5JBVO3PXB7OH6L/graph.json","events_json":"https://pith.science/api/pith-number/42LIIEDICQJM5JBVO3PXB7OH6L/events.json","paper":"https://pith.science/paper/42LIIEDI"},"agent_actions":{"view_html":"https://pith.science/pith/42LIIEDICQJM5JBVO3PXB7OH6L","download_json":"https://pith.science/pith/42LIIEDICQJM5JBVO3PXB7OH6L.json","view_paper":"https://pith.science/paper/42LIIEDI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2511.15684&json=true","fetch_graph":"https://pith.science/api/pith-number/42LIIEDICQJM5JBVO3PXB7OH6L/graph.json","fetch_events":"https://pith.science/api/pith-number/42LIIEDICQJM5JBVO3PXB7OH6L/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/42LIIEDICQJM5JBVO3PXB7OH6L/action/timestamp_anchor","attest_storage":"https://pith.science/pith/42LIIEDICQJM5JBVO3PXB7OH6L/action/storage_attestation","attest_author":"https://pith.science/pith/42LIIEDICQJM5JBVO3PXB7OH6L/action/author_attestation","sign_citation":"https://pith.science/pith/42LIIEDICQJM5JBVO3PXB7OH6L/action/citation_signature","submit_replication":"https://pith.science/pith/42LIIEDICQJM5JBVO3PXB7OH6L/action/replication_record"}},"created_at":"2026-07-07T02:18:30.461247+00:00","updated_at":"2026-07-07T02:18:30.461247+00:00"}