{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:S6TQ2AMIRN27S3GZCDQHXOJAUG","short_pith_number":"pith:S6TQ2AMI","schema_version":"1.0","canonical_sha256":"97a70d01888b75f96cd910e07bb920a1a21e82619d23a7c5599ba09e509367de","source":{"kind":"arxiv","id":"2505.17998","version":1},"attestation_state":"computed","paper":{"title":"TRACE for Tracking the Emergence of Semantic Representations in Transformers","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Andr\\'e Freitas, Danilo S. Carvalho, Nura Aljaafari","submitted_at":"2025-05-23T15:03:51Z","abstract_excerpt":"Modern transformer models exhibit phase transitions during training, distinct shifts from memorisation to abstraction, but the mechanisms underlying these transitions remain poorly understood. Prior work has often focused on endpoint representations or isolated signals like curvature or mutual information, typically in symbolic or arithmetic domains, overlooking the emergence of linguistic structure. We introduce TRACE (Tracking Representation Abstraction and Compositional Emergence), a diagnostic framework combining geometric, informational, and linguistic signals to detect phase transitions "},"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":"2505.17998","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-23T15:03:51Z","cross_cats_sorted":[],"title_canon_sha256":"f75510bc3993fddc801c49e7c71b3eda53e384888bb51588e2762b98ec77689b","abstract_canon_sha256":"5e129404feeed0ee69dfc16306b81420bbd02edced2eeb2a2a7e30a4a1b6f3c5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:08:35.000683Z","signature_b64":"CLmKGk/Ypqg+ybMmWorKDk0kmfUB5sS/VI1ArA0TqBznPj+JXj10XYcEIp3J+HVkt0kR6GePSjKiPuB2NIE8Bg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"97a70d01888b75f96cd910e07bb920a1a21e82619d23a7c5599ba09e509367de","last_reissued_at":"2026-07-05T11:08:35.000155Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:08:35.000155Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"TRACE for Tracking the Emergence of Semantic Representations in Transformers","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Andr\\'e Freitas, Danilo S. Carvalho, Nura Aljaafari","submitted_at":"2025-05-23T15:03:51Z","abstract_excerpt":"Modern transformer models exhibit phase transitions during training, distinct shifts from memorisation to abstraction, but the mechanisms underlying these transitions remain poorly understood. Prior work has often focused on endpoint representations or isolated signals like curvature or mutual information, typically in symbolic or arithmetic domains, overlooking the emergence of linguistic structure. We introduce TRACE (Tracking Representation Abstraction and Compositional Emergence), a diagnostic framework combining geometric, informational, and linguistic signals to detect phase transitions "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.17998","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/2505.17998/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":"2505.17998","created_at":"2026-07-05T11:08:35.000211+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.17998v1","created_at":"2026-07-05T11:08:35.000211+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.17998","created_at":"2026-07-05T11:08:35.000211+00:00"},{"alias_kind":"pith_short_12","alias_value":"S6TQ2AMIRN27","created_at":"2026-07-05T11:08:35.000211+00:00"},{"alias_kind":"pith_short_16","alias_value":"S6TQ2AMIRN27S3GZ","created_at":"2026-07-05T11:08:35.000211+00:00"},{"alias_kind":"pith_short_8","alias_value":"S6TQ2AMI","created_at":"2026-07-05T11:08:35.000211+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.20276","citing_title":"Rethinking Intrinsic Dimension Estimation in Neural Representations","ref_index":27,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/S6TQ2AMIRN27S3GZCDQHXOJAUG","json":"https://pith.science/pith/S6TQ2AMIRN27S3GZCDQHXOJAUG.json","graph_json":"https://pith.science/api/pith-number/S6TQ2AMIRN27S3GZCDQHXOJAUG/graph.json","events_json":"https://pith.science/api/pith-number/S6TQ2AMIRN27S3GZCDQHXOJAUG/events.json","paper":"https://pith.science/paper/S6TQ2AMI"},"agent_actions":{"view_html":"https://pith.science/pith/S6TQ2AMIRN27S3GZCDQHXOJAUG","download_json":"https://pith.science/pith/S6TQ2AMIRN27S3GZCDQHXOJAUG.json","view_paper":"https://pith.science/paper/S6TQ2AMI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.17998&json=true","fetch_graph":"https://pith.science/api/pith-number/S6TQ2AMIRN27S3GZCDQHXOJAUG/graph.json","fetch_events":"https://pith.science/api/pith-number/S6TQ2AMIRN27S3GZCDQHXOJAUG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/S6TQ2AMIRN27S3GZCDQHXOJAUG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/S6TQ2AMIRN27S3GZCDQHXOJAUG/action/storage_attestation","attest_author":"https://pith.science/pith/S6TQ2AMIRN27S3GZCDQHXOJAUG/action/author_attestation","sign_citation":"https://pith.science/pith/S6TQ2AMIRN27S3GZCDQHXOJAUG/action/citation_signature","submit_replication":"https://pith.science/pith/S6TQ2AMIRN27S3GZCDQHXOJAUG/action/replication_record"}},"created_at":"2026-07-05T11:08:35.000211+00:00","updated_at":"2026-07-05T11:08:35.000211+00:00"}