{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:I5PBLHCISM53LBVGUS5CSQOQBD","short_pith_number":"pith:I5PBLHCI","schema_version":"1.0","canonical_sha256":"475e159c48933bb586a6a4ba2941d008d3e5607e0a538af8f65035d327d310a6","source":{"kind":"arxiv","id":"2402.02366","version":2},"attestation_state":"computed","paper":{"title":"Transolver: A Fast Transformer Solver for PDEs on General Geometries","license":"http://creativecommons.org/licenses/by/4.0/","headline":"By grouping mesh points with similar physical states into learnable slices, Transolver lets Transformers solve PDEs on arbitrary geometries in linear time.","cross_cats":["cs.NA","math.NA"],"primary_cat":"cs.LG","authors_text":"Haixu Wu, Haowen Wang, Huakun Luo, Jianmin Wang, Mingsheng Long","submitted_at":"2024-02-04T06:37:38Z","abstract_excerpt":"Transformers have empowered many milestones across various fields and have recently been applied to solve partial differential equations (PDEs). However, since PDEs are typically discretized into large-scale meshes with complex geometries, it is challenging for Transformers to capture intricate physical correlations directly from massive individual points. Going beyond superficial and unwieldy meshes, we present Transolver based on a more foundational idea, which is learning intrinsic physical states hidden behind discretized geometries. Specifically, we propose a new Physics-Attention to adap"},"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":true,"formal_links_present":true},"canonical_record":{"source":{"id":"2402.02366","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-02-04T06:37:38Z","cross_cats_sorted":["cs.NA","math.NA"],"title_canon_sha256":"1d220a973c657e8af81c81d5fd504f3a1c2281406a4885737ee3ee7d591b8003","abstract_canon_sha256":"7bc5f6994e368c87d7d5131fb94f441ea35873ed905b7481f22811602b775829"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:38:50.065851Z","signature_b64":"Wq3fI5TvANfDoo8EcDqwwOayLBFVVhaiDrHSXKcE9FFrGs/mC9c/H5MKnUIHvIvajXX9dOgOd6fiYSXwxiJiDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"475e159c48933bb586a6a4ba2941d008d3e5607e0a538af8f65035d327d310a6","last_reissued_at":"2026-05-17T23:38:50.065373Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:38:50.065373Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Transolver: A Fast Transformer Solver for PDEs on General Geometries","license":"http://creativecommons.org/licenses/by/4.0/","headline":"By grouping mesh points with similar physical states into learnable slices, Transolver lets Transformers solve PDEs on arbitrary geometries in linear time.","cross_cats":["cs.NA","math.NA"],"primary_cat":"cs.LG","authors_text":"Haixu Wu, Haowen Wang, Huakun Luo, Jianmin Wang, Mingsheng Long","submitted_at":"2024-02-04T06:37:38Z","abstract_excerpt":"Transformers have empowered many milestones across various fields and have recently been applied to solve partial differential equations (PDEs). However, since PDEs are typically discretized into large-scale meshes with complex geometries, it is challenging for Transformers to capture intricate physical correlations directly from massive individual points. Going beyond superficial and unwieldy meshes, we present Transolver based on a more foundational idea, which is learning intrinsic physical states hidden behind discretized geometries. Specifically, we propose a new Physics-Attention to adap"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Transolver achieves consistent state-of-the-art with 22% relative gain across six standard benchmarks and also excels in large-scale industrial simulations, including car and airfoil designs.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That mesh points under similar physical states can be reliably and adaptively grouped into a series of learnable slices whose encoded tokens capture the necessary physical correlations without missing critical local interactions.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Transolver learns intrinsic physical states from discretized meshes by adaptively splitting domains into flexible learnable slices and computing attention over physics-aware tokens, achieving state-of-the-art PDE solving on general geometries.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"By grouping mesh points with similar physical states into learnable slices, Transolver lets Transformers solve PDEs on arbitrary geometries in linear time.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"340733dc3c2fca53346349ebdddfcf8973164eff6a26d45370eb4fe7393fa51a"},"source":{"id":"2402.02366","kind":"arxiv","version":2},"verdict":{"id":"cd55237e-63ee-4d4a-acf3-b36c1c8fdbba","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T21:37:13.966590Z","strongest_claim":"Transolver achieves consistent state-of-the-art with 22% relative gain across six standard benchmarks and also excels in large-scale industrial simulations, including car and airfoil designs.","one_line_summary":"Transolver learns intrinsic physical states from discretized meshes by adaptively splitting domains into flexible learnable slices and computing attention over physics-aware tokens, achieving state-of-the-art PDE solving on general geometries.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That mesh points under similar physical states can be reliably and adaptively grouped into a series of learnable slices whose encoded tokens capture the necessary physical correlations without missing critical local interactions.","pith_extraction_headline":"By grouping mesh points with similar physical states into learnable slices, Transolver lets Transformers solve PDEs on arbitrary geometries in linear time."},"references":{"count":33,"sample":[{"doi":"","year":null,"title":"GPT-4 Technical Report","work_id":"b928e041-6991-4c08-8c81-0359e4097c7b","ref_index":1,"cited_arxiv_id":"2303.08774","is_internal_anchor":true},{"doi":"","year":null,"title":"ShapeNet: An Information-Rich 3D Model Repository","work_id":"b2ac5b60-daa9-435b-9369-12271e126edd","ref_index":2,"cited_arxiv_id":"1512.03012","is_internal_anchor":true},{"doi":"","year":null,"title":"Training Deep Nets with Sublinear Memory Cost","work_id":"f2c5c287-a500-40e4-a136-e7e3172db1d7","ref_index":3,"cited_arxiv_id":"1604.06174","is_internal_anchor":true},{"doi":"","year":2003,"title":"Neural Operator: Graph Kernel Network for Partial Differential Equations","work_id":"00a591bc-6cad-477c-be12-26d5623f625d","ref_index":4,"cited_arxiv_id":"2003.03485","is_internal_anchor":true},{"doi":"","year":2003,"title":"Geometry-informed neural operator for large-scale 3d pdes","work_id":"b57f5179-2379-4048-b180-5c84d9a20a8a","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":33,"snapshot_sha256":"d12429c3edc1b9404e29e68c2b2eb9802467a2ace0fd478295ad4c00a30f2b0f","internal_anchors":4},"formal_canon":{"evidence_count":2,"snapshot_sha256":"9279b3a01e583a3d2fda6c36ba3669370ed74f26cabc678872ee6c9d1b9ad58e"},"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":"2402.02366","created_at":"2026-05-17T23:38:50.065463+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.02366v2","created_at":"2026-05-17T23:38:50.065463+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.02366","created_at":"2026-05-17T23:38:50.065463+00:00"},{"alias_kind":"pith_short_12","alias_value":"I5PBLHCISM53","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_16","alias_value":"I5PBLHCISM53LBVG","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_8","alias_value":"I5PBLHCI","created_at":"2026-05-18T12:33:37.589309+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":24,"internal_anchor_count":24,"sample":[{"citing_arxiv_id":"2605.19565","citing_title":"HiLiftAeroML: High-Fidelity Computational Fluid Dynamics Dataset for High-Lift Aircraft Aerodynamics","ref_index":13,"is_internal_anchor":true},{"citing_arxiv_id":"2605.03548","citing_title":"PerFlow: Physics-Embedded Rectified Flow for Efficient Reconstruction and Uncertainty Quantification of Spatiotemporal Dynamics","ref_index":34,"is_internal_anchor":true},{"citing_arxiv_id":"2605.15231","citing_title":"Mask-Morph Graph U-Net: A Generalisable Mesh-Based Surrogate for Crashworthiness Field Prediction under Large Geometric Variation","ref_index":22,"is_internal_anchor":true},{"citing_arxiv_id":"2512.14397","citing_title":"SuperWing: a comprehensive transonic wing dataset for data-driven aerodynamic design","ref_index":26,"is_internal_anchor":true},{"citing_arxiv_id":"2602.15472","citing_title":"Fluids You Can Trust: Property-Preserving Operator Learning for Incompressible Flows","ref_index":13,"is_internal_anchor":true},{"citing_arxiv_id":"2605.14546","citing_title":"Discovering Physical Directions in Weight Space: Composing Neural PDE Experts","ref_index":34,"is_internal_anchor":true},{"citing_arxiv_id":"2605.08935","citing_title":"PnP-Corrector: A Universal Correction Framework for Coupled Spatiotemporal Forecasting","ref_index":156,"is_internal_anchor":true},{"citing_arxiv_id":"2605.12965","citing_title":"U-HNO: A U-shaped Hybrid Neural Operator with Sparse-Point Adaptive Routing for Non-stationary PDE Dynamics","ref_index":30,"is_internal_anchor":true},{"citing_arxiv_id":"2605.13790","citing_title":"Di-BiLPS: Denoising induced Bidirectional Latent-PDE-Solver under Sparse Observations","ref_index":16,"is_internal_anchor":true},{"citing_arxiv_id":"2605.11111","citing_title":"ShardTensor: Domain Parallelism for Scientific Machine Learning","ref_index":38,"is_internal_anchor":true},{"citing_arxiv_id":"2605.08935","citing_title":"PnP-Corrector: A Universal Correction Framework for Coupled Spatiotemporal Forecasting","ref_index":156,"is_internal_anchor":true},{"citing_arxiv_id":"2605.10451","citing_title":"Don't Fix the Basis -- Learn It: Spectral Representation with Adaptive Basis Learning for PDEs","ref_index":25,"is_internal_anchor":true},{"citing_arxiv_id":"2605.08436","citing_title":"A meshfree exterior calculus for generalizable and data-efficient learning of physics from point clouds","ref_index":23,"is_internal_anchor":true},{"citing_arxiv_id":"2605.09016","citing_title":"CATO: Charted Attention for Neural PDE Operators","ref_index":30,"is_internal_anchor":true},{"citing_arxiv_id":"2604.25985","citing_title":"Learning Neural Operator Surrogates for the Black Hole Accretion Code","ref_index":43,"is_internal_anchor":true},{"citing_arxiv_id":"2605.03548","citing_title":"PerFlow: Physics-Embedded Rectified Flow for Efficient Reconstruction and Uncertainty Quantification of Spatiotemporal Dynamics","ref_index":32,"is_internal_anchor":true},{"citing_arxiv_id":"2605.04832","citing_title":"Replay-Based Continual Learning for Physics-Informed Neural Operators","ref_index":7,"is_internal_anchor":true},{"citing_arxiv_id":"2605.05586","citing_title":"AeroJEPA: Learning Semantic Latent Representations for Scalable 3D Aerodynamic Field Modeling","ref_index":37,"is_internal_anchor":true},{"citing_arxiv_id":"2605.00062","citing_title":"RETO: A Rotary-Enhanced Transformer Operator for High-Fidelity Prediction of Automotive Aerodynamics","ref_index":10,"is_internal_anchor":true},{"citing_arxiv_id":"2605.07098","citing_title":"CarCrashNet: A Large-Scale Dataset and Hierarchical Neural Solver for Data-Driven Structural Crash Simulation","ref_index":126,"is_internal_anchor":true},{"citing_arxiv_id":"2604.15617","citing_title":"A Structure-Preserving Graph Neural Solver for Parametric Hyperbolic Conservation Laws","ref_index":32,"is_internal_anchor":true},{"citing_arxiv_id":"2604.16649","citing_title":"FLARE: A Data-Efficient Surrogate for Predicting Displacement Fields in Directed Energy Deposition","ref_index":36,"is_internal_anchor":true},{"citing_arxiv_id":"2604.18491","citing_title":"Faster by Design: Interactive Aerodynamics via Neural Surrogates Trained on Expert-Validated CFD","ref_index":12,"is_internal_anchor":true},{"citing_arxiv_id":"2604.21101","citing_title":"A Hybridizable Neural Time Integrator for Stable Autoregressive Forecasting","ref_index":40,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":2,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/I5PBLHCISM53LBVGUS5CSQOQBD","json":"https://pith.science/pith/I5PBLHCISM53LBVGUS5CSQOQBD.json","graph_json":"https://pith.science/api/pith-number/I5PBLHCISM53LBVGUS5CSQOQBD/graph.json","events_json":"https://pith.science/api/pith-number/I5PBLHCISM53LBVGUS5CSQOQBD/events.json","paper":"https://pith.science/paper/I5PBLHCI"},"agent_actions":{"view_html":"https://pith.science/pith/I5PBLHCISM53LBVGUS5CSQOQBD","download_json":"https://pith.science/pith/I5PBLHCISM53LBVGUS5CSQOQBD.json","view_paper":"https://pith.science/paper/I5PBLHCI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.02366&json=true","fetch_graph":"https://pith.science/api/pith-number/I5PBLHCISM53LBVGUS5CSQOQBD/graph.json","fetch_events":"https://pith.science/api/pith-number/I5PBLHCISM53LBVGUS5CSQOQBD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/I5PBLHCISM53LBVGUS5CSQOQBD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/I5PBLHCISM53LBVGUS5CSQOQBD/action/storage_attestation","attest_author":"https://pith.science/pith/I5PBLHCISM53LBVGUS5CSQOQBD/action/author_attestation","sign_citation":"https://pith.science/pith/I5PBLHCISM53LBVGUS5CSQOQBD/action/citation_signature","submit_replication":"https://pith.science/pith/I5PBLHCISM53LBVGUS5CSQOQBD/action/replication_record"}},"created_at":"2026-05-17T23:38:50.065463+00:00","updated_at":"2026-05-17T23:38:50.065463+00:00"}