{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:6RPRGXPQQ2X2VOC6TDFVUSJAMV","short_pith_number":"pith:6RPRGXPQ","schema_version":"1.0","canonical_sha256":"f45f135df086afaab85e98cb5a49206553fbea71bbf39221e74268b2c0e59120","source":{"kind":"arxiv","id":"2109.01050","version":2},"attestation_state":"computed","paper":{"title":"Characterizing possible failure modes in physics-informed neural networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.NA","math.NA","physics.comp-ph"],"primary_cat":"cs.LG","authors_text":"Aditi S. Krishnapriyan, Amir Gholami, Michael W. Mahoney, Robert M. Kirby, Shandian Zhe","submitted_at":"2021-09-02T16:06:45Z","abstract_excerpt":"Recent work in scientific machine learning has developed so-called physics-informed neural network (PINN) models. The typical approach is to incorporate physical domain knowledge as soft constraints on an empirical loss function and use existing machine learning methodologies to train the model. We demonstrate that, while existing PINN methodologies can learn good models for relatively trivial problems, they can easily fail to learn relevant physical phenomena for even slightly more complex problems. In particular, we analyze several distinct situations of widespread physical interest, includi"},"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":"2109.01050","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-09-02T16:06:45Z","cross_cats_sorted":["cs.AI","cs.NA","math.NA","physics.comp-ph"],"title_canon_sha256":"49b182d75a61e4ea6e7682c21993068c205c6568f23aeeaff3be4bcac988c9ad","abstract_canon_sha256":"4a3bc65d96172d1e5f1d8c9af6b6d80b690dbaded05b2ecb445762325236c0f4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:30:53.891659Z","signature_b64":"CfC1hWzEuRllN02IBScN8qZ7mHUPmvgvcYDkgLqCdPHKbh8T0BEUtRD59iXxXUB9fYxJA9GlKoNiLM6/EVFrCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f45f135df086afaab85e98cb5a49206553fbea71bbf39221e74268b2c0e59120","last_reissued_at":"2026-07-05T03:30:53.891192Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:30:53.891192Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Characterizing possible failure modes in physics-informed neural networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.NA","math.NA","physics.comp-ph"],"primary_cat":"cs.LG","authors_text":"Aditi S. Krishnapriyan, Amir Gholami, Michael W. Mahoney, Robert M. Kirby, Shandian Zhe","submitted_at":"2021-09-02T16:06:45Z","abstract_excerpt":"Recent work in scientific machine learning has developed so-called physics-informed neural network (PINN) models. The typical approach is to incorporate physical domain knowledge as soft constraints on an empirical loss function and use existing machine learning methodologies to train the model. We demonstrate that, while existing PINN methodologies can learn good models for relatively trivial problems, they can easily fail to learn relevant physical phenomena for even slightly more complex problems. In particular, we analyze several distinct situations of widespread physical interest, includi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2109.01050","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/2109.01050/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":"2109.01050","created_at":"2026-07-05T03:30:53.891247+00:00"},{"alias_kind":"arxiv_version","alias_value":"2109.01050v2","created_at":"2026-07-05T03:30:53.891247+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2109.01050","created_at":"2026-07-05T03:30:53.891247+00:00"},{"alias_kind":"pith_short_12","alias_value":"6RPRGXPQQ2X2","created_at":"2026-07-05T03:30:53.891247+00:00"},{"alias_kind":"pith_short_16","alias_value":"6RPRGXPQQ2X2VOC6","created_at":"2026-07-05T03:30:53.891247+00:00"},{"alias_kind":"pith_short_8","alias_value":"6RPRGXPQ","created_at":"2026-07-05T03:30:53.891247+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":8,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.18175","citing_title":"A Convex Quasilinearization Method for Solving Nonlinear PDEs with Physics-Informed Neural Networks","ref_index":28,"is_internal_anchor":false},{"citing_arxiv_id":"2606.26128","citing_title":"Physics-guided Convolutional Neural Network for Domain Growth Prediction in Systems with Conserved Kinetics","ref_index":36,"is_internal_anchor":false},{"citing_arxiv_id":"2606.02335","citing_title":"Neural Spectral Element Methods for stiff multiphysics PDEs with electrochemical transport benchmarks","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2603.29184","citing_title":"Cell-induced densification and tether formation in fibrous extracellular matrices with biomimetic physics-informed neural networks","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2604.02438","citing_title":"Mitigating Data Scarcity in Spaceflight Applications for Offline Reinforcement Learning Using Physics-Informed Deep Generative Models","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2604.23528","citing_title":"When PINNs Go Wrong: Pseudo-Time Stepping Against Spurious Solutions","ref_index":46,"is_internal_anchor":false},{"citing_arxiv_id":"2605.00510","citing_title":"Scale-Aware Adversarial Analysis: A Diagnostic for Generative AI in Multiscale Complex Systems","ref_index":76,"is_internal_anchor":false},{"citing_arxiv_id":"2605.00308","citing_title":"Adaptive anisotropic composite quadratures for residual minimisation in neural PDE approximations","ref_index":27,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6RPRGXPQQ2X2VOC6TDFVUSJAMV","json":"https://pith.science/pith/6RPRGXPQQ2X2VOC6TDFVUSJAMV.json","graph_json":"https://pith.science/api/pith-number/6RPRGXPQQ2X2VOC6TDFVUSJAMV/graph.json","events_json":"https://pith.science/api/pith-number/6RPRGXPQQ2X2VOC6TDFVUSJAMV/events.json","paper":"https://pith.science/paper/6RPRGXPQ"},"agent_actions":{"view_html":"https://pith.science/pith/6RPRGXPQQ2X2VOC6TDFVUSJAMV","download_json":"https://pith.science/pith/6RPRGXPQQ2X2VOC6TDFVUSJAMV.json","view_paper":"https://pith.science/paper/6RPRGXPQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2109.01050&json=true","fetch_graph":"https://pith.science/api/pith-number/6RPRGXPQQ2X2VOC6TDFVUSJAMV/graph.json","fetch_events":"https://pith.science/api/pith-number/6RPRGXPQQ2X2VOC6TDFVUSJAMV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6RPRGXPQQ2X2VOC6TDFVUSJAMV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6RPRGXPQQ2X2VOC6TDFVUSJAMV/action/storage_attestation","attest_author":"https://pith.science/pith/6RPRGXPQQ2X2VOC6TDFVUSJAMV/action/author_attestation","sign_citation":"https://pith.science/pith/6RPRGXPQQ2X2VOC6TDFVUSJAMV/action/citation_signature","submit_replication":"https://pith.science/pith/6RPRGXPQQ2X2VOC6TDFVUSJAMV/action/replication_record"}},"created_at":"2026-07-05T03:30:53.891247+00:00","updated_at":"2026-07-05T03:30:53.891247+00:00"}