{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:BGA2EXWQ7X6HCSKX2DECNUZNXU","short_pith_number":"pith:BGA2EXWQ","schema_version":"1.0","canonical_sha256":"0981a25ed0fdfc714957d0c826d32dbd0ba449dfcdc374b1919be23be66c8297","source":{"kind":"arxiv","id":"2504.06231","version":2},"attestation_state":"computed","paper":{"title":"Orb-v3: atomistic simulation at scale","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cond-mat.mtrl-sci","authors_text":"Benjamin Rhodes, James Gin, Jonathan Godwin, Mark Neumann, Sander Vandenhaute, Tim Duignan, Vaidotas \\v{S}imkus","submitted_at":"2025-04-08T17:27:34Z","abstract_excerpt":"We introduce Orb-v3, the next generation of the Orb family of universal interatomic potentials. Models in this family expand the performance-speed-memory Pareto frontier, offering near SoTA performance across a range of evaluations with a >10x reduction in latency and > 8x reduction in memory. Our experiments systematically traverse this frontier, charting the trade-off induced by roto-equivariance, conservatism and graph sparsity. Contrary to recent literature, we find that non-equivariant, non-conservative architectures can accurately model physical properties, including those which require "},"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":"2504.06231","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cond-mat.mtrl-sci","submitted_at":"2025-04-08T17:27:34Z","cross_cats_sorted":[],"title_canon_sha256":"00feeba0d15703488e250230429b6e337d7357c12417202ced0472269b168d81","abstract_canon_sha256":"5fbbb9c8d879d9c0cf6415ae50ac6c437f71fc48ebf68755c962428923058fa5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:47:39.132484Z","signature_b64":"twmuyo5pyN6gFN5Vcxv56FR0/QTN1qTXS6WZ51RO7j5tDUoRHHaazlpD8NatwJrH7W3suo7lIfMt1VSEIWDhAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0981a25ed0fdfc714957d0c826d32dbd0ba449dfcdc374b1919be23be66c8297","last_reissued_at":"2026-07-05T10:47:39.132016Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:47:39.132016Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Orb-v3: atomistic simulation at scale","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cond-mat.mtrl-sci","authors_text":"Benjamin Rhodes, James Gin, Jonathan Godwin, Mark Neumann, Sander Vandenhaute, Tim Duignan, Vaidotas \\v{S}imkus","submitted_at":"2025-04-08T17:27:34Z","abstract_excerpt":"We introduce Orb-v3, the next generation of the Orb family of universal interatomic potentials. Models in this family expand the performance-speed-memory Pareto frontier, offering near SoTA performance across a range of evaluations with a >10x reduction in latency and > 8x reduction in memory. Our experiments systematically traverse this frontier, charting the trade-off induced by roto-equivariance, conservatism and graph sparsity. Contrary to recent literature, we find that non-equivariant, non-conservative architectures can accurately model physical properties, including those which require "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2504.06231","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/2504.06231/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":"2504.06231","created_at":"2026-07-05T10:47:39.132070+00:00"},{"alias_kind":"arxiv_version","alias_value":"2504.06231v2","created_at":"2026-07-05T10:47:39.132070+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2504.06231","created_at":"2026-07-05T10:47:39.132070+00:00"},{"alias_kind":"pith_short_12","alias_value":"BGA2EXWQ7X6H","created_at":"2026-07-05T10:47:39.132070+00:00"},{"alias_kind":"pith_short_16","alias_value":"BGA2EXWQ7X6HCSKX","created_at":"2026-07-05T10:47:39.132070+00:00"},{"alias_kind":"pith_short_8","alias_value":"BGA2EXWQ","created_at":"2026-07-05T10:47:39.132070+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":30,"internal_anchor_count":3,"sample":[{"citing_arxiv_id":"2607.08224","citing_title":"Phase stability and ionic transport in post-spinel CaV$_2$O$_4$ cathode","ref_index":60,"is_internal_anchor":true},{"citing_arxiv_id":"2607.07606","citing_title":"MLIP Studio: An Open Platform for Interactive Benchmarking and Atomistic Simulations Using Machine Learning Interatomic Potentials","ref_index":17,"is_internal_anchor":true},{"citing_arxiv_id":"2607.07647","citing_title":"Are Machine Learning Interatomic Potentials Truly Practical? A Benchmark of 23 Mainstream Models","ref_index":4,"is_internal_anchor":true},{"citing_arxiv_id":"2607.01362","citing_title":"Enerzyme: A Framework for Efficient Training of Reactive Neural Network Potentials for Enzyme Catalysis with Application to Methyltransferases","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2607.01713","citing_title":"Predicting Novel Stable Materials for Experimental Synthesis","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2607.02236","citing_title":"Efficient Large-Scale STEM-EELS Simulations With Torched-TACAW","ref_index":41,"is_internal_anchor":false},{"citing_arxiv_id":"2606.11072","citing_title":"Approaching the Limit of Intrinsic Crystalline Thermal Insulation","ref_index":50,"is_internal_anchor":false},{"citing_arxiv_id":"2606.07327","citing_title":"Six Open Questions in Machine-Learned Interatomic Potential Foundation Models","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2606.04728","citing_title":"Curvature-driven revival of charge density waves in non-Euclidean space","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2606.02455","citing_title":"Speculative Sampling For Faster Molecular Dynamics","ref_index":62,"is_internal_anchor":false},{"citing_arxiv_id":"2605.19939","citing_title":"Uncertainty-aware Machine Learning Interatomic Potentials via Learned Functional Perturbations","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2605.27640","citing_title":"Additive binding energies in asphalt on a quantum processor via quantum-selected configuration interaction (QSCI)","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2605.29029","citing_title":"Geometry-based Discovery of Calcium Battery Cathodes Accelerated by Foundational Machine-Learned Models","ref_index":70,"is_internal_anchor":false},{"citing_arxiv_id":"2606.23214","citing_title":"Universal Interatomic Potentials as Configuration-Space Generators for One-Shot and Iterative Fine-Tuning of Ab Initio-Accurate Material-Specific Models","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2605.16612","citing_title":"PRISMat: Policy-Driven, Permutation-Invariant Autoregressive Material Generation","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2605.19939","citing_title":"Uncertainty-aware Machine Learning Interatomic Potentials via Learned Functional Perturbations","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15630","citing_title":"Reweighting free energy profiles between universal machine learning interatomic potentials for fast consensus building","ref_index":64,"is_internal_anchor":false},{"citing_arxiv_id":"2605.13788","citing_title":"Force-Aware Neural Tangent Kernels for Scalable and Robust Active Learning of MLIPs","ref_index":41,"is_internal_anchor":false},{"citing_arxiv_id":"2605.03964","citing_title":"Pretrained Model Representations as Acquisition Signals for Active Learning of MLIPs","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2512.05717","citing_title":"Comparing the latent features of universal machine-learning interatomic potentials","ref_index":65,"is_internal_anchor":false},{"citing_arxiv_id":"2601.16331","citing_title":"Accuracy and Efficiency Benchmarks of Pretrained Machine Learning Potentials for Molecular Simulations","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2605.14527","citing_title":"Lang2MLIP: End-to-End Language-to-Machine Learning Interatomic Potential Development with Autonomous Agentic Workflows","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2605.13788","citing_title":"Force-Aware Neural Tangent Kernels for Scalable and Robust Active Learning of MLIPs","ref_index":36,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11610","citing_title":"Fast and Accurate Prediction of Lattice Thermal Conductivity via Machine Learning Surrogates","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08262","citing_title":"SLayerGen: a Crystal Generative Model for all Space and Layer Groups","ref_index":86,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/BGA2EXWQ7X6HCSKX2DECNUZNXU","json":"https://pith.science/pith/BGA2EXWQ7X6HCSKX2DECNUZNXU.json","graph_json":"https://pith.science/api/pith-number/BGA2EXWQ7X6HCSKX2DECNUZNXU/graph.json","events_json":"https://pith.science/api/pith-number/BGA2EXWQ7X6HCSKX2DECNUZNXU/events.json","paper":"https://pith.science/paper/BGA2EXWQ"},"agent_actions":{"view_html":"https://pith.science/pith/BGA2EXWQ7X6HCSKX2DECNUZNXU","download_json":"https://pith.science/pith/BGA2EXWQ7X6HCSKX2DECNUZNXU.json","view_paper":"https://pith.science/paper/BGA2EXWQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2504.06231&json=true","fetch_graph":"https://pith.science/api/pith-number/BGA2EXWQ7X6HCSKX2DECNUZNXU/graph.json","fetch_events":"https://pith.science/api/pith-number/BGA2EXWQ7X6HCSKX2DECNUZNXU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BGA2EXWQ7X6HCSKX2DECNUZNXU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BGA2EXWQ7X6HCSKX2DECNUZNXU/action/storage_attestation","attest_author":"https://pith.science/pith/BGA2EXWQ7X6HCSKX2DECNUZNXU/action/author_attestation","sign_citation":"https://pith.science/pith/BGA2EXWQ7X6HCSKX2DECNUZNXU/action/citation_signature","submit_replication":"https://pith.science/pith/BGA2EXWQ7X6HCSKX2DECNUZNXU/action/replication_record"}},"created_at":"2026-07-05T10:47:39.132070+00:00","updated_at":"2026-07-05T10:47:39.132070+00:00"}