{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:52V7YDRL63VQ5SRTIZCGOKBGAV","short_pith_number":"pith:52V7YDRL","schema_version":"1.0","canonical_sha256":"eeabfc0e2bf6eb0eca334644672826055b17eba1f69f67c9b263eea47d71872f","source":{"kind":"arxiv","id":"2605.21060","version":1},"attestation_state":"computed","paper":{"title":"Divide et Calibra: Multiclass Local Calibration via Vector Quantization","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Andrea Passerini, Andrea Pugnana, Cesare Barbera, Giovanni De Toni, Lorenzo Perini","submitted_at":"2026-05-20T11:44:55Z","abstract_excerpt":"Accurate and well-calibrated Machine Learning (ML) models are mandatory in high-stakes settings, yet effective multiclass calibration remains challenging: global approaches assume calibration errors are homogeneous across the latent space, while local methods often rely on latent-space dimensionality reduction, which leads to information loss. To address these issues, we propose a compositional approach to multiclass calibration, where region-specific calibration maps are constructed from shared codeword-dependent factors. We instantiate this idea via Vector Quantization (VQ), which induces a "},"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":"2605.21060","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-20T11:44:55Z","cross_cats_sorted":["cs.AI","stat.ML"],"title_canon_sha256":"035de884ae44939fa6b1975a57a75936368daacc0854d43b47cc623ae854e41d","abstract_canon_sha256":"02ea82ed29a8abeef8a187545ab057fa30ca8776108fa0fe09a316dbc300b776"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-21T01:05:34.612245Z","signature_b64":"M2sPRhAB9y0l1sX4RtIFBaS1X0JzwjcJz9f2xFDi+Nyf+aOaBGWgENS6x0/+l+jIuQUZtooDK9iT1NnfNd4YDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"eeabfc0e2bf6eb0eca334644672826055b17eba1f69f67c9b263eea47d71872f","last_reissued_at":"2026-05-21T01:05:34.611506Z","signature_status":"signed_v1","first_computed_at":"2026-05-21T01:05:34.611506Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Divide et Calibra: Multiclass Local Calibration via Vector Quantization","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Andrea Passerini, Andrea Pugnana, Cesare Barbera, Giovanni De Toni, Lorenzo Perini","submitted_at":"2026-05-20T11:44:55Z","abstract_excerpt":"Accurate and well-calibrated Machine Learning (ML) models are mandatory in high-stakes settings, yet effective multiclass calibration remains challenging: global approaches assume calibration errors are homogeneous across the latent space, while local methods often rely on latent-space dimensionality reduction, which leads to information loss. To address these issues, we propose a compositional approach to multiclass calibration, where region-specific calibration maps are constructed from shared codeword-dependent factors. We instantiate this idea via Vector Quantization (VQ), which induces a "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.21060","kind":"arxiv","version":1},"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/2605.21060/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":"2605.21060","created_at":"2026-05-21T01:05:34.611627+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.21060v1","created_at":"2026-05-21T01:05:34.611627+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.21060","created_at":"2026-05-21T01:05:34.611627+00:00"},{"alias_kind":"pith_short_12","alias_value":"52V7YDRL63VQ","created_at":"2026-05-21T01:05:34.611627+00:00"},{"alias_kind":"pith_short_16","alias_value":"52V7YDRL63VQ5SRT","created_at":"2026-05-21T01:05:34.611627+00:00"},{"alias_kind":"pith_short_8","alias_value":"52V7YDRL","created_at":"2026-05-21T01:05:34.611627+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/52V7YDRL63VQ5SRTIZCGOKBGAV","json":"https://pith.science/pith/52V7YDRL63VQ5SRTIZCGOKBGAV.json","graph_json":"https://pith.science/api/pith-number/52V7YDRL63VQ5SRTIZCGOKBGAV/graph.json","events_json":"https://pith.science/api/pith-number/52V7YDRL63VQ5SRTIZCGOKBGAV/events.json","paper":"https://pith.science/paper/52V7YDRL"},"agent_actions":{"view_html":"https://pith.science/pith/52V7YDRL63VQ5SRTIZCGOKBGAV","download_json":"https://pith.science/pith/52V7YDRL63VQ5SRTIZCGOKBGAV.json","view_paper":"https://pith.science/paper/52V7YDRL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.21060&json=true","fetch_graph":"https://pith.science/api/pith-number/52V7YDRL63VQ5SRTIZCGOKBGAV/graph.json","fetch_events":"https://pith.science/api/pith-number/52V7YDRL63VQ5SRTIZCGOKBGAV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/52V7YDRL63VQ5SRTIZCGOKBGAV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/52V7YDRL63VQ5SRTIZCGOKBGAV/action/storage_attestation","attest_author":"https://pith.science/pith/52V7YDRL63VQ5SRTIZCGOKBGAV/action/author_attestation","sign_citation":"https://pith.science/pith/52V7YDRL63VQ5SRTIZCGOKBGAV/action/citation_signature","submit_replication":"https://pith.science/pith/52V7YDRL63VQ5SRTIZCGOKBGAV/action/replication_record"}},"created_at":"2026-05-21T01:05:34.611627+00:00","updated_at":"2026-05-21T01:05:34.611627+00:00"}