{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:7ECUX2WT3FVS2IHK4OZBKP6AL6","short_pith_number":"pith:7ECUX2WT","canonical_record":{"source":{"id":"2604.23354","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"eess.AS","submitted_at":"2026-04-25T15:44:20Z","cross_cats_sorted":["cs.AI","eess.SP"],"title_canon_sha256":"ae83f33cba4c15379fe40b1661e10e8ff8fdb173de9ca753b34d66df75b37c45","abstract_canon_sha256":"b846fd5324a4735d8795f09020c794678597c233ede5167750785e06b97572d1"},"schema_version":"1.0"},"canonical_sha256":"f9054bead3d96b2d20eae3b2153fc05f965498ef02a3c8ce2306eff1901e3ba4","source":{"kind":"arxiv","id":"2604.23354","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2604.23354","created_at":"2026-05-29T01:05:10Z"},{"alias_kind":"arxiv_version","alias_value":"2604.23354v2","created_at":"2026-05-29T01:05:10Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2604.23354","created_at":"2026-05-29T01:05:10Z"},{"alias_kind":"pith_short_12","alias_value":"7ECUX2WT3FVS","created_at":"2026-05-29T01:05:10Z"},{"alias_kind":"pith_short_16","alias_value":"7ECUX2WT3FVS2IHK","created_at":"2026-05-29T01:05:10Z"},{"alias_kind":"pith_short_8","alias_value":"7ECUX2WT","created_at":"2026-05-29T01:05:10Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:7ECUX2WT3FVS2IHK4OZBKP6AL6","target":"record","payload":{"canonical_record":{"source":{"id":"2604.23354","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"eess.AS","submitted_at":"2026-04-25T15:44:20Z","cross_cats_sorted":["cs.AI","eess.SP"],"title_canon_sha256":"ae83f33cba4c15379fe40b1661e10e8ff8fdb173de9ca753b34d66df75b37c45","abstract_canon_sha256":"b846fd5324a4735d8795f09020c794678597c233ede5167750785e06b97572d1"},"schema_version":"1.0"},"canonical_sha256":"f9054bead3d96b2d20eae3b2153fc05f965498ef02a3c8ce2306eff1901e3ba4","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-29T01:05:10.515376Z","signature_b64":"2xQg1PIfYOamGNIaSsEn0wazw9p62TTHyVKBA0LufUyICub+fpnvk8ujlKC95hCl3JLoizLu+NBMBIuxDf6hDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f9054bead3d96b2d20eae3b2153fc05f965498ef02a3c8ce2306eff1901e3ba4","last_reissued_at":"2026-05-29T01:05:10.514741Z","signature_status":"signed_v1","first_computed_at":"2026-05-29T01:05:10.514741Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2604.23354","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-29T01:05:10Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"GHPiZ8QjVYxoQ1tKY1H3L1CCyhzeE2pUZNK4SwIJmkxrp9NwQU6y0idxuxm/rfFSjG5bl6lpGRB4XadMM/6UCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-10T01:14:36.518459Z"},"content_sha256":"7b4984cbd9dfda8af78c118d23ebeb15ca05f6368d02e53521f29e5fc3356d75","schema_version":"1.0","event_id":"sha256:7b4984cbd9dfda8af78c118d23ebeb15ca05f6368d02e53521f29e5fc3356d75"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:7ECUX2WT3FVS2IHK4OZBKP6AL6","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Explainable AI in Speaker Recognition -- Making Latent Representations Understandable","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Speaker recognition neural networks organize their latent representations into hierarchical clusters that align with semantic attributes like gender and nationality.","cross_cats":["cs.AI","eess.SP"],"primary_cat":"eess.AS","authors_text":"Mark D. Plumbley, Wenwu Wang, Yanze Xu","submitted_at":"2026-04-25T15:44:20Z","abstract_excerpt":"Neural networks can be trained to learn task-relevant representations from data. Understanding how these networks make decisions falls within the Explainable AI (XAI) domain. This paper proposes to study an XAI topic: uncovering the unknown organisation in the representations, particularly those a speaker recognition network learns from utterances, for recognising speaker identity.\n  Past studies have employed algorithms (e.g. K-means) to analyse how network representations can be naturally organised into independent clusters in different ways, i.e., to analyse flat clustering phenomena within"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"This work applies SLINK and HDBSCAN to demonstrate the existence of hierarchical clustering phenomena within the network representation space, and designs HCCM to perform one-to-one matching between predefined semantic classes and hierarchical representation clusters, with Liebig's score to quantify performance.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the hierarchical clusters produced by SLINK or HDBSCAN correspond to meaningful semantic classes or their conjunctions in a non-arbitrary way that HCCM can reliably detect and that Liebig's score meaningfully diagnoses limiting factors.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Speaker recognition networks form hierarchical clusters in latent space that can be matched to semantic classes using new HCCM algorithm and quantified by Liebig's score.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Speaker recognition neural networks organize their latent representations into hierarchical clusters that align with semantic attributes like gender and nationality.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"95ac01f5a4f2f342730ed2bffdd8e2c841b50b8c74678370b22c7286c4506168"},"source":{"id":"2604.23354","kind":"arxiv","version":2},"verdict":{"id":"d070c761-f6c7-4423-95bd-7990dfc25d6d","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-08T06:54:00.456162Z","strongest_claim":"This work applies SLINK and HDBSCAN to demonstrate the existence of hierarchical clustering phenomena within the network representation space, and designs HCCM to perform one-to-one matching between predefined semantic classes and hierarchical representation clusters, with Liebig's score to quantify performance.","one_line_summary":"Speaker recognition networks form hierarchical clusters in latent space that can be matched to semantic classes using new HCCM algorithm and quantified by Liebig's score.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the hierarchical clusters produced by SLINK or HDBSCAN correspond to meaningful semantic classes or their conjunctions in a non-arbitrary way that HCCM can reliably detect and that Liebig's score meaningfully diagnoses limiting factors.","pith_extraction_headline":"Speaker recognition neural networks organize their latent representations into hierarchical clusters that align with semantic attributes like gender and nationality."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.23354/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"ai_meta_artifact","ran_at":"2026-05-21T09:34:00.617979Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T23:14:59.591565Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"9ccaa68e2e110782a580f1553f549b074e112222139037f3ce5d24641a8f5993"},"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"},"verdict_id":"d070c761-f6c7-4423-95bd-7990dfc25d6d"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-29T01:05:10Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"/u5iLvNiZ+wFDi57gy4Vw+POcnYkHazEtF/d+TrBlLSjtYz8zAG+TOOc2u3R04Oia+c3ZI/2S2CB/CwxWhY2Dg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-10T01:14:36.519517Z"},"content_sha256":"e2465f7b0350639bfd984f31192e2e53b900a57c0b5b24b96d7f3a3a508602c1","schema_version":"1.0","event_id":"sha256:e2465f7b0350639bfd984f31192e2e53b900a57c0b5b24b96d7f3a3a508602c1"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/7ECUX2WT3FVS2IHK4OZBKP6AL6/bundle.json","state_url":"https://pith.science/pith/7ECUX2WT3FVS2IHK4OZBKP6AL6/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/7ECUX2WT3FVS2IHK4OZBKP6AL6/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":"2026-06-10T01:14:36Z","links":{"resolver":"https://pith.science/pith/7ECUX2WT3FVS2IHK4OZBKP6AL6","bundle":"https://pith.science/pith/7ECUX2WT3FVS2IHK4OZBKP6AL6/bundle.json","state":"https://pith.science/pith/7ECUX2WT3FVS2IHK4OZBKP6AL6/state.json","well_known_bundle":"https://pith.science/.well-known/pith/7ECUX2WT3FVS2IHK4OZBKP6AL6/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:7ECUX2WT3FVS2IHK4OZBKP6AL6","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":"b846fd5324a4735d8795f09020c794678597c233ede5167750785e06b97572d1","cross_cats_sorted":["cs.AI","eess.SP"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"eess.AS","submitted_at":"2026-04-25T15:44:20Z","title_canon_sha256":"ae83f33cba4c15379fe40b1661e10e8ff8fdb173de9ca753b34d66df75b37c45"},"schema_version":"1.0","source":{"id":"2604.23354","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2604.23354","created_at":"2026-05-29T01:05:10Z"},{"alias_kind":"arxiv_version","alias_value":"2604.23354v2","created_at":"2026-05-29T01:05:10Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2604.23354","created_at":"2026-05-29T01:05:10Z"},{"alias_kind":"pith_short_12","alias_value":"7ECUX2WT3FVS","created_at":"2026-05-29T01:05:10Z"},{"alias_kind":"pith_short_16","alias_value":"7ECUX2WT3FVS2IHK","created_at":"2026-05-29T01:05:10Z"},{"alias_kind":"pith_short_8","alias_value":"7ECUX2WT","created_at":"2026-05-29T01:05:10Z"}],"graph_snapshots":[{"event_id":"sha256:e2465f7b0350639bfd984f31192e2e53b900a57c0b5b24b96d7f3a3a508602c1","target":"graph","created_at":"2026-05-29T01:05:10Z","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":"This work applies SLINK and HDBSCAN to demonstrate the existence of hierarchical clustering phenomena within the network representation space, and designs HCCM to perform one-to-one matching between predefined semantic classes and hierarchical representation clusters, with Liebig's score to quantify performance."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That the hierarchical clusters produced by SLINK or HDBSCAN correspond to meaningful semantic classes or their conjunctions in a non-arbitrary way that HCCM can reliably detect and that Liebig's score meaningfully diagnoses limiting factors."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"Speaker recognition networks form hierarchical clusters in latent space that can be matched to semantic classes using new HCCM algorithm and quantified by Liebig's score."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Speaker recognition neural networks organize their latent representations into hierarchical clusters that align with semantic attributes like gender and nationality."}],"snapshot_sha256":"95ac01f5a4f2f342730ed2bffdd8e2c841b50b8c74678370b22c7286c4506168"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[{"findings_count":0,"name":"ai_meta_artifact","ran_at":"2026-05-21T09:34:00.617979Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"doi_compliance","ran_at":"2026-05-19T23:14:59.591565Z","status":"completed","version":"1.0.0"}],"endpoint":"/pith/2604.23354/integrity.json","findings":[],"snapshot_sha256":"9ccaa68e2e110782a580f1553f549b074e112222139037f3ce5d24641a8f5993","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Neural networks can be trained to learn task-relevant representations from data. Understanding how these networks make decisions falls within the Explainable AI (XAI) domain. This paper proposes to study an XAI topic: uncovering the unknown organisation in the representations, particularly those a speaker recognition network learns from utterances, for recognising speaker identity.\n  Past studies have employed algorithms (e.g. K-means) to analyse how network representations can be naturally organised into independent clusters in different ways, i.e., to analyse flat clustering phenomena within","authors_text":"Mark D. Plumbley, Wenwu Wang, Yanze Xu","cross_cats":["cs.AI","eess.SP"],"headline":"Speaker recognition neural networks organize their latent representations into hierarchical clusters that align with semantic attributes like gender and nationality.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"eess.AS","submitted_at":"2026-04-25T15:44:20Z","title":"Explainable AI in Speaker Recognition -- Making Latent Representations Understandable"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2604.23354","kind":"arxiv","version":2},"verdict":{"created_at":"2026-05-08T06:54:00.456162Z","id":"d070c761-f6c7-4423-95bd-7990dfc25d6d","model_set":{"reader":"grok-4.3"},"one_line_summary":"Speaker recognition networks form hierarchical clusters in latent space that can be matched to semantic classes using new HCCM algorithm and quantified by Liebig's score.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Speaker recognition neural networks organize their latent representations into hierarchical clusters that align with semantic attributes like gender and nationality.","strongest_claim":"This work applies SLINK and HDBSCAN to demonstrate the existence of hierarchical clustering phenomena within the network representation space, and designs HCCM to perform one-to-one matching between predefined semantic classes and hierarchical representation clusters, with Liebig's score to quantify performance.","weakest_assumption":"That the hierarchical clusters produced by SLINK or HDBSCAN correspond to meaningful semantic classes or their conjunctions in a non-arbitrary way that HCCM can reliably detect and that Liebig's score meaningfully diagnoses limiting factors."}},"verdict_id":"d070c761-f6c7-4423-95bd-7990dfc25d6d"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:7b4984cbd9dfda8af78c118d23ebeb15ca05f6368d02e53521f29e5fc3356d75","target":"record","created_at":"2026-05-29T01:05:10Z","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":"b846fd5324a4735d8795f09020c794678597c233ede5167750785e06b97572d1","cross_cats_sorted":["cs.AI","eess.SP"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"eess.AS","submitted_at":"2026-04-25T15:44:20Z","title_canon_sha256":"ae83f33cba4c15379fe40b1661e10e8ff8fdb173de9ca753b34d66df75b37c45"},"schema_version":"1.0","source":{"id":"2604.23354","kind":"arxiv","version":2}},"canonical_sha256":"f9054bead3d96b2d20eae3b2153fc05f965498ef02a3c8ce2306eff1901e3ba4","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f9054bead3d96b2d20eae3b2153fc05f965498ef02a3c8ce2306eff1901e3ba4","first_computed_at":"2026-05-29T01:05:10.514741Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-29T01:05:10.514741Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"2xQg1PIfYOamGNIaSsEn0wazw9p62TTHyVKBA0LufUyICub+fpnvk8ujlKC95hCl3JLoizLu+NBMBIuxDf6hDA==","signature_status":"signed_v1","signed_at":"2026-05-29T01:05:10.515376Z","signed_message":"canonical_sha256_bytes"},"source_id":"2604.23354","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:7b4984cbd9dfda8af78c118d23ebeb15ca05f6368d02e53521f29e5fc3356d75","sha256:e2465f7b0350639bfd984f31192e2e53b900a57c0b5b24b96d7f3a3a508602c1"],"state_sha256":"f0542de8ec66cae1c47ba92f75442947406b59684a67a1f389422c990bd99e0b"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"j3kiCVEK2aLAySYzg8HQeFxlkgE4PYxwcrbwlOY+s1iuI+7IuM5wY3zQvhrAh7Ha1CkQTIXsmpOBoGiI7T4KDw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-10T01:14:36.524568Z","bundle_sha256":"7f999044a5155b6912ea1880c34b0be727f1aa1e2643dda0c92559c456346a24"}}