{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:4X2OE44CQT6NZBCKWTSLATHQXY","short_pith_number":"pith:4X2OE44C","canonical_record":{"source":{"id":"2509.07024","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"physics.plasm-ph","submitted_at":"2025-09-07T09:36:51Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"07b7416266227bdabb642c12b230892d905f2f2febb632e756b8545e6e20bd7a","abstract_canon_sha256":"a0b9c1ae3905c3cf97d694a016cd748d1cfdb0bd79100dac3c3ee7b6713021d6"},"schema_version":"1.0"},"canonical_sha256":"e5f4e2738284fcdc844ab4e4b04cf0be011500ee1aae9be1a6042c6784d5af81","source":{"kind":"arxiv","id":"2509.07024","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2509.07024","created_at":"2026-05-20T01:04:57Z"},{"alias_kind":"arxiv_version","alias_value":"2509.07024v3","created_at":"2026-05-20T01:04:57Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.07024","created_at":"2026-05-20T01:04:57Z"},{"alias_kind":"pith_short_12","alias_value":"4X2OE44CQT6N","created_at":"2026-05-20T01:04:57Z"},{"alias_kind":"pith_short_16","alias_value":"4X2OE44CQT6NZBCK","created_at":"2026-05-20T01:04:57Z"},{"alias_kind":"pith_short_8","alias_value":"4X2OE44C","created_at":"2026-05-20T01:04:57Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:4X2OE44CQT6NZBCKWTSLATHQXY","target":"record","payload":{"canonical_record":{"source":{"id":"2509.07024","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"physics.plasm-ph","submitted_at":"2025-09-07T09:36:51Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"07b7416266227bdabb642c12b230892d905f2f2febb632e756b8545e6e20bd7a","abstract_canon_sha256":"a0b9c1ae3905c3cf97d694a016cd748d1cfdb0bd79100dac3c3ee7b6713021d6"},"schema_version":"1.0"},"canonical_sha256":"e5f4e2738284fcdc844ab4e4b04cf0be011500ee1aae9be1a6042c6784d5af81","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T01:04:57.106348Z","signature_b64":"CoXY/OMQK/ujJ9/worK7itBHiGqNg7/J5mG9dj2q9msLeTh63vaFJWs0EC0deWx+1uTnEVRzMtXUTEyAJJEhBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e5f4e2738284fcdc844ab4e4b04cf0be011500ee1aae9be1a6042c6784d5af81","last_reissued_at":"2026-05-20T01:04:57.105558Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T01:04:57.105558Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2509.07024","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-20T01:04:57Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"F0mPBXqG5WGQ3akewc05QicPTppcJ/RBo8cKQwKHN+zZ5Z9B5FOOQKN8EocBOrxljd0wc1Rv939hZwDDFP1pDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-01T21:34:04.793244Z"},"content_sha256":"7d8dab9f652935b7b1103d8b66ddbb28bc097807cc3e9bfc2aad344f5753e3de","schema_version":"1.0","event_id":"sha256:7d8dab9f652935b7b1103d8b66ddbb28bc097807cc3e9bfc2aad344f5753e3de"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:4X2OE44CQT6NZBCKWTSLATHQXY","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"TGLF-WINN: Data-Efficient Deep Learning Surrogate for Turbulent Transport Modeling in Fusion","license":"http://creativecommons.org/licenses/by/4.0/","headline":"TGLF-WINN achieves full accuracy of turbulent transport predictions using only 25% of the training data through Bayesian active learning and physics-guided regularization.","cross_cats":["cs.LG"],"primary_cat":"physics.plasm-ph","authors_text":"Brian Sammuli, Futian Zhang, Lawson Fuller, Orso Meneghini, Raffi Nazikian, Rose Yu, Sterling Smith, Tom Neiser, Wesley Liu, Yadi Cao","submitted_at":"2025-09-07T09:36:51Z","abstract_excerpt":"The Trapped Gyro-Landau Fluid (TGLF) model provides fast, accurate predictions of turbulent transport in tokamaks, but whole device simulations requiring thousands of evaluations remain computationally expensive. Neural network (NN) surrogates offer accelerated inference with fully differentiable approximations that enable gradient-based coupling but typically require large training datasets to capture transport flux variations across plasma conditions, creating significant training burden and limiting applicability to expensive gyrokinetic simulations. We propose TGLF-WINN (Wavenumber-Informe"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Adding Bayesian Active Learning, TGLF-WINN matches TGLF-NN's full-data offline accuracy using only 25% of the training data, within 2.8% of TGLF-NN's full-data baseline and 4.3% of our own full-data result.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The training dataset (even when subsampled) is statistically representative of the full range of plasma conditions encountered in whole-device simulations, so that uncertainty estimates from the Bayesian model reliably guide selection of informative points without missing important regimes.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"TGLF-WINN matches standard neural network surrogate accuracy for TGLF transport fluxes using only 25% of the training data via feature tuning, physics-guided wavenumber regularization, and Bayesian active learning.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"TGLF-WINN achieves full accuracy of turbulent transport predictions using only 25% of the training data through Bayesian active learning and physics-guided regularization.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"8f83174a145b2f235ca2a052fc2ffbb3ce71f6b86a28395c791089913626122b"},"source":{"id":"2509.07024","kind":"arxiv","version":3},"verdict":{"id":"7340f55b-a932-4987-8095-d460cfdc8b33","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-18T18:27:54.910394Z","strongest_claim":"Adding Bayesian Active Learning, TGLF-WINN matches TGLF-NN's full-data offline accuracy using only 25% of the training data, within 2.8% of TGLF-NN's full-data baseline and 4.3% of our own full-data result.","one_line_summary":"TGLF-WINN matches standard neural network surrogate accuracy for TGLF transport fluxes using only 25% of the training data via feature tuning, physics-guided wavenumber regularization, and Bayesian active learning.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The training dataset (even when subsampled) is statistically representative of the full range of plasma conditions encountered in whole-device simulations, so that uncertainty estimates from the Bayesian model reliably guide selection of informative points without missing important regimes.","pith_extraction_headline":"TGLF-WINN achieves full accuracy of turbulent transport predictions using only 25% of the training data through Bayesian active learning and physics-guided regularization."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2509.07024/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":2,"snapshot_sha256":"f7f0cbfb3ea6362c210747f90cb7d3c8f44b993f69ca2dee5bf249cd87fbc590"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"7340f55b-a932-4987-8095-d460cfdc8b33"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-20T01:04:57Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"lJIPANUr631y1hVfYat+frzKLVaU0wxJVkmAz+cH5wIvAnE0xV8HCPj30aLBjwT21WjZRq0BRCHglaB1/carAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-01T21:34:04.793736Z"},"content_sha256":"87339f93fbe894376b203a63a678064fdf6c9e088c4e2a029aa44e82ff2535ec","schema_version":"1.0","event_id":"sha256:87339f93fbe894376b203a63a678064fdf6c9e088c4e2a029aa44e82ff2535ec"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/4X2OE44CQT6NZBCKWTSLATHQXY/bundle.json","state_url":"https://pith.science/pith/4X2OE44CQT6NZBCKWTSLATHQXY/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/4X2OE44CQT6NZBCKWTSLATHQXY/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-06-01T21:34:04Z","links":{"resolver":"https://pith.science/pith/4X2OE44CQT6NZBCKWTSLATHQXY","bundle":"https://pith.science/pith/4X2OE44CQT6NZBCKWTSLATHQXY/bundle.json","state":"https://pith.science/pith/4X2OE44CQT6NZBCKWTSLATHQXY/state.json","well_known_bundle":"https://pith.science/.well-known/pith/4X2OE44CQT6NZBCKWTSLATHQXY/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:4X2OE44CQT6NZBCKWTSLATHQXY","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"a0b9c1ae3905c3cf97d694a016cd748d1cfdb0bd79100dac3c3ee7b6713021d6","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"physics.plasm-ph","submitted_at":"2025-09-07T09:36:51Z","title_canon_sha256":"07b7416266227bdabb642c12b230892d905f2f2febb632e756b8545e6e20bd7a"},"schema_version":"1.0","source":{"id":"2509.07024","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2509.07024","created_at":"2026-05-20T01:04:57Z"},{"alias_kind":"arxiv_version","alias_value":"2509.07024v3","created_at":"2026-05-20T01:04:57Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.07024","created_at":"2026-05-20T01:04:57Z"},{"alias_kind":"pith_short_12","alias_value":"4X2OE44CQT6N","created_at":"2026-05-20T01:04:57Z"},{"alias_kind":"pith_short_16","alias_value":"4X2OE44CQT6NZBCK","created_at":"2026-05-20T01:04:57Z"},{"alias_kind":"pith_short_8","alias_value":"4X2OE44C","created_at":"2026-05-20T01:04:57Z"}],"graph_snapshots":[{"event_id":"sha256:87339f93fbe894376b203a63a678064fdf6c9e088c4e2a029aa44e82ff2535ec","target":"graph","created_at":"2026-05-20T01:04:57Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"Adding Bayesian Active Learning, TGLF-WINN matches TGLF-NN's full-data offline accuracy using only 25% of the training data, within 2.8% of TGLF-NN's full-data baseline and 4.3% of our own full-data result."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"The training dataset (even when subsampled) is statistically representative of the full range of plasma conditions encountered in whole-device simulations, so that uncertainty estimates from the Bayesian model reliably guide selection of informative points without missing important regimes."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"TGLF-WINN matches standard neural network surrogate accuracy for TGLF transport fluxes using only 25% of the training data via feature tuning, physics-guided wavenumber regularization, and Bayesian active learning."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"TGLF-WINN achieves full accuracy of turbulent transport predictions using only 25% of the training data through Bayesian active learning and physics-guided regularization."}],"snapshot_sha256":"8f83174a145b2f235ca2a052fc2ffbb3ce71f6b86a28395c791089913626122b"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"f7f0cbfb3ea6362c210747f90cb7d3c8f44b993f69ca2dee5bf249cd87fbc590"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2509.07024/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The Trapped Gyro-Landau Fluid (TGLF) model provides fast, accurate predictions of turbulent transport in tokamaks, but whole device simulations requiring thousands of evaluations remain computationally expensive. Neural network (NN) surrogates offer accelerated inference with fully differentiable approximations that enable gradient-based coupling but typically require large training datasets to capture transport flux variations across plasma conditions, creating significant training burden and limiting applicability to expensive gyrokinetic simulations. We propose TGLF-WINN (Wavenumber-Informe","authors_text":"Brian Sammuli, Futian Zhang, Lawson Fuller, Orso Meneghini, Raffi Nazikian, Rose Yu, Sterling Smith, Tom Neiser, Wesley Liu, Yadi Cao","cross_cats":["cs.LG"],"headline":"TGLF-WINN achieves full accuracy of turbulent transport predictions using only 25% of the training data through Bayesian active learning and physics-guided regularization.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"physics.plasm-ph","submitted_at":"2025-09-07T09:36:51Z","title":"TGLF-WINN: Data-Efficient Deep Learning Surrogate for Turbulent Transport Modeling in Fusion"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2509.07024","kind":"arxiv","version":3},"verdict":{"created_at":"2026-05-18T18:27:54.910394Z","id":"7340f55b-a932-4987-8095-d460cfdc8b33","model_set":{"reader":"grok-4.3"},"one_line_summary":"TGLF-WINN matches standard neural network surrogate accuracy for TGLF transport fluxes using only 25% of the training data via feature tuning, physics-guided wavenumber regularization, and Bayesian active learning.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"TGLF-WINN achieves full accuracy of turbulent transport predictions using only 25% of the training data through Bayesian active learning and physics-guided regularization.","strongest_claim":"Adding Bayesian Active Learning, TGLF-WINN matches TGLF-NN's full-data offline accuracy using only 25% of the training data, within 2.8% of TGLF-NN's full-data baseline and 4.3% of our own full-data result.","weakest_assumption":"The training dataset (even when subsampled) is statistically representative of the full range of plasma conditions encountered in whole-device simulations, so that uncertainty estimates from the Bayesian model reliably guide selection of informative points without missing important regimes."}},"verdict_id":"7340f55b-a932-4987-8095-d460cfdc8b33"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:7d8dab9f652935b7b1103d8b66ddbb28bc097807cc3e9bfc2aad344f5753e3de","target":"record","created_at":"2026-05-20T01:04:57Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"a0b9c1ae3905c3cf97d694a016cd748d1cfdb0bd79100dac3c3ee7b6713021d6","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"physics.plasm-ph","submitted_at":"2025-09-07T09:36:51Z","title_canon_sha256":"07b7416266227bdabb642c12b230892d905f2f2febb632e756b8545e6e20bd7a"},"schema_version":"1.0","source":{"id":"2509.07024","kind":"arxiv","version":3}},"canonical_sha256":"e5f4e2738284fcdc844ab4e4b04cf0be011500ee1aae9be1a6042c6784d5af81","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"e5f4e2738284fcdc844ab4e4b04cf0be011500ee1aae9be1a6042c6784d5af81","first_computed_at":"2026-05-20T01:04:57.105558Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-20T01:04:57.105558Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"CoXY/OMQK/ujJ9/worK7itBHiGqNg7/J5mG9dj2q9msLeTh63vaFJWs0EC0deWx+1uTnEVRzMtXUTEyAJJEhBA==","signature_status":"signed_v1","signed_at":"2026-05-20T01:04:57.106348Z","signed_message":"canonical_sha256_bytes"},"source_id":"2509.07024","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:7d8dab9f652935b7b1103d8b66ddbb28bc097807cc3e9bfc2aad344f5753e3de","sha256:87339f93fbe894376b203a63a678064fdf6c9e088c4e2a029aa44e82ff2535ec"],"state_sha256":"28b4cb411f0a5aa7898d2a73ca63d3e33816273f943918b6ac0fd6091c7701b3"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"nhojp33WSVDMIk/+/JCDQoFsfjjJRA6exdDXpDtn248sccQf5YsdpJMYdELPZQaImlDpMIP2EarG/7pkyGQoAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-01T21:34:04.796096Z","bundle_sha256":"4db3fc165e404ba518225abac48f4a0dd79154a6ecf52a7e6eb446281b90420b"}}