{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:7ZBIHT5IPJ3CRPZ4PIYBLERMEP","short_pith_number":"pith:7ZBIHT5I","schema_version":"1.0","canonical_sha256":"fe4283cfa87a7628bf3c7a3015922c23d507ea885be200feef5c5cc299028cb8","source":{"kind":"arxiv","id":"2512.07420","version":2},"attestation_state":"computed","paper":{"title":"E-PCN: Jet Tagging with Explainable Particle Chebyshev Networks Using Kinematic Features","license":"http://creativecommons.org/licenses/by/4.0/","headline":"E-PCN classifies jets by building four graphs each weighted by a different kinematic variable and uses Grad-CAM to show angular separation plus transverse momentum drive 76 percent of decisions.","cross_cats":["cs.LG","hep-ex"],"primary_cat":"hep-ph","authors_text":"Adrita Khan, AKM Mahbubur Rahman, Amin Ahsan Ali, Choudhury Ben Yamin Siddiqui, M. Arshad Momen, Md Raqibul Islam, Md. Zakir Hossan, Mir Sazzat Hossain, Tanjib Khan","submitted_at":"2025-12-08T10:53:05Z","abstract_excerpt":"The identification and classification of collimated particle sprays, or jets, are essential for interpreting data from high-energy collider experiments. While deep learning has improved jet classification, it often lacks interpretability. We introduce the Explainable Particle Chebyshev Network (E-PCN), a graph neural network extending the Particle Chebyshev Network (PCN). E-PCN integrates kinematic variables into jet classification by constructing four graph representations per jet, each weighted by a distinct variable: angular separation ($\\Delta$), transverse momentum ($k_T$), momentum fract"},"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":"2512.07420","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"hep-ph","submitted_at":"2025-12-08T10:53:05Z","cross_cats_sorted":["cs.LG","hep-ex"],"title_canon_sha256":"764c76a6fd007d048c0822e905fc534da5938263774f16ce47854adb248ec187","abstract_canon_sha256":"43ab21b14fd439f59d49c5d6f72d3e33c5f123494caf36bd8ec33b3770b54ce6"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-21T01:04:19.881880Z","signature_b64":"BedNnlDa23jDhQA4oOQWMru3PfBHF6mj9XD25KGXbInMQHXfwzz3AePT54piAXlcBXBhYsvv8YjF4fZOAVQAAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fe4283cfa87a7628bf3c7a3015922c23d507ea885be200feef5c5cc299028cb8","last_reissued_at":"2026-05-21T01:04:19.880953Z","signature_status":"signed_v1","first_computed_at":"2026-05-21T01:04:19.880953Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"E-PCN: Jet Tagging with Explainable Particle Chebyshev Networks Using Kinematic Features","license":"http://creativecommons.org/licenses/by/4.0/","headline":"E-PCN classifies jets by building four graphs each weighted by a different kinematic variable and uses Grad-CAM to show angular separation plus transverse momentum drive 76 percent of decisions.","cross_cats":["cs.LG","hep-ex"],"primary_cat":"hep-ph","authors_text":"Adrita Khan, AKM Mahbubur Rahman, Amin Ahsan Ali, Choudhury Ben Yamin Siddiqui, M. Arshad Momen, Md Raqibul Islam, Md. Zakir Hossan, Mir Sazzat Hossain, Tanjib Khan","submitted_at":"2025-12-08T10:53:05Z","abstract_excerpt":"The identification and classification of collimated particle sprays, or jets, are essential for interpreting data from high-energy collider experiments. While deep learning has improved jet classification, it often lacks interpretability. We introduce the Explainable Particle Chebyshev Network (E-PCN), a graph neural network extending the Particle Chebyshev Network (PCN). E-PCN integrates kinematic variables into jet classification by constructing four graph representations per jet, each weighted by a distinct variable: angular separation ($\\Delta$), transverse momentum ($k_T$), momentum fract"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Evaluated on the JetClass dataset with 10 signal classes, E-PCN achieves a macro-accuracy of 94.67%, macro-AUC of 96.78%, and macro-AUPR of 86.79%, representing improvements of 2.36%, 4.13%, and 24.88% respectively over the baseline PCN implementation, while demonstrating physically interpretable feature learning with angular separation and transverse momentum accounting for approximately 76% of classification decisions.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That Grad-CAM attributions on the four kinematic-weighted graphs faithfully reflect the true causal importance of those variables in the model's internal decision process rather than explanation artifacts.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"E-PCN reaches 94.67% macro-accuracy on 10-class jet tagging by weighting graphs with angular separation, transverse momentum, momentum fraction, and invariant mass, with Grad-CAM showing the first two account for 76% of decisions and yielding gains over baseline PCN.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"E-PCN classifies jets by building four graphs each weighted by a different kinematic variable and uses Grad-CAM to show angular separation plus transverse momentum drive 76 percent of decisions.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"5f73cc273124dbfe0a52a9e329655cd69492b13753c0ed832b4488a99c5cff90"},"source":{"id":"2512.07420","kind":"arxiv","version":2},"verdict":{"id":"b1ce2acf-9651-44ff-ac3a-f15df72b6d2e","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-17T01:01:55.638906Z","strongest_claim":"Evaluated on the JetClass dataset with 10 signal classes, E-PCN achieves a macro-accuracy of 94.67%, macro-AUC of 96.78%, and macro-AUPR of 86.79%, representing improvements of 2.36%, 4.13%, and 24.88% respectively over the baseline PCN implementation, while demonstrating physically interpretable feature learning with angular separation and transverse momentum accounting for approximately 76% of classification decisions.","one_line_summary":"E-PCN reaches 94.67% macro-accuracy on 10-class jet tagging by weighting graphs with angular separation, transverse momentum, momentum fraction, and invariant mass, with Grad-CAM showing the first two account for 76% of decisions and yielding gains over baseline PCN.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That Grad-CAM attributions on the four kinematic-weighted graphs faithfully reflect the true causal importance of those variables in the model's internal decision process rather than explanation artifacts.","pith_extraction_headline":"E-PCN classifies jets by building four graphs each weighted by a different kinematic variable and uses Grad-CAM to show angular separation plus transverse momentum drive 76 percent of decisions."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2512.07420/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":51,"sample":[{"doi":"10.23731/cyrm-2020-0010","year":2020,"title":"CERN Yellow Reports: Monographs","work_id":"0f3b270b-32a5-424b-b3a4-e7858e9ef186","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2018,"title":"High-Luminosity LHC","work_id":"5182c324-3e93-4354-b0be-a20cc02c7be5","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2024,"title":"S. Mondal and L. Mastrolorenzo,Machine learning in high energy physics: a review of heavy-flavor jet tagging at the LHC,Eur. Phys. J. ST233(2024) 2657 [arXiv:2404.01071] [inSPIRE]. – 16 –","work_id":"5265ec18-3d68-4b68-9d75-73038bcca244","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2018,"title":"Machine Learning in High Energy Physics Community White Paper","work_id":"e3284bfb-3295-4e13-bcff-02cf9f2eec49","ref_index":4,"cited_arxiv_id":"1807.02876","is_internal_anchor":true},{"doi":"","year":2020,"title":"H. Qu and L. Gouskos,ParticleNet: Jet Tagging via Particle Clouds,Phys. Rev. D101 (2020) 056019 [arXiv:1902.08570] [inSPIRE]","work_id":"ff4ad651-a8a7-49cf-9113-279cc5097adf","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":51,"snapshot_sha256":"e3852a356982b6c47a37feaa06cca86b5de3a64c40b9ec0d1f8116b9671a9837","internal_anchors":14},"formal_canon":{"evidence_count":2,"snapshot_sha256":"ac786ff5f17e6e51de4e3a8345236cfb11eca1e831dba88725523179d27047ee"},"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":"2512.07420","created_at":"2026-05-21T01:04:19.881105+00:00"},{"alias_kind":"arxiv_version","alias_value":"2512.07420v2","created_at":"2026-05-21T01:04:19.881105+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2512.07420","created_at":"2026-05-21T01:04:19.881105+00:00"},{"alias_kind":"pith_short_12","alias_value":"7ZBIHT5IPJ3C","created_at":"2026-05-21T01:04:19.881105+00:00"},{"alias_kind":"pith_short_16","alias_value":"7ZBIHT5IPJ3CRPZ4","created_at":"2026-05-21T01:04:19.881105+00:00"},{"alias_kind":"pith_short_8","alias_value":"7ZBIHT5I","created_at":"2026-05-21T01:04:19.881105+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2604.25885","citing_title":"Explainable AI for Jet Tagging: A Comparative Study of GNNExplainer, GNNShap, and GradCAM for Jet Tagging in the Lund Jet Plane","ref_index":46,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":2,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/7ZBIHT5IPJ3CRPZ4PIYBLERMEP","json":"https://pith.science/pith/7ZBIHT5IPJ3CRPZ4PIYBLERMEP.json","graph_json":"https://pith.science/api/pith-number/7ZBIHT5IPJ3CRPZ4PIYBLERMEP/graph.json","events_json":"https://pith.science/api/pith-number/7ZBIHT5IPJ3CRPZ4PIYBLERMEP/events.json","paper":"https://pith.science/paper/7ZBIHT5I"},"agent_actions":{"view_html":"https://pith.science/pith/7ZBIHT5IPJ3CRPZ4PIYBLERMEP","download_json":"https://pith.science/pith/7ZBIHT5IPJ3CRPZ4PIYBLERMEP.json","view_paper":"https://pith.science/paper/7ZBIHT5I","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2512.07420&json=true","fetch_graph":"https://pith.science/api/pith-number/7ZBIHT5IPJ3CRPZ4PIYBLERMEP/graph.json","fetch_events":"https://pith.science/api/pith-number/7ZBIHT5IPJ3CRPZ4PIYBLERMEP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7ZBIHT5IPJ3CRPZ4PIYBLERMEP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7ZBIHT5IPJ3CRPZ4PIYBLERMEP/action/storage_attestation","attest_author":"https://pith.science/pith/7ZBIHT5IPJ3CRPZ4PIYBLERMEP/action/author_attestation","sign_citation":"https://pith.science/pith/7ZBIHT5IPJ3CRPZ4PIYBLERMEP/action/citation_signature","submit_replication":"https://pith.science/pith/7ZBIHT5IPJ3CRPZ4PIYBLERMEP/action/replication_record"}},"created_at":"2026-05-21T01:04:19.881105+00:00","updated_at":"2026-05-21T01:04:19.881105+00:00"}