{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:G3KSXZ5NMHFGZXD7VSURZAE42T","short_pith_number":"pith:G3KSXZ5N","schema_version":"1.0","canonical_sha256":"36d52be7ad61ca6cdc7faca91c809cd4cc109c3110791b51a24c53c3ce7bb0b3","source":{"kind":"arxiv","id":"2604.24374","version":2},"attestation_state":"computed","paper":{"title":"MIPIC: Matryoshka Representation Learning via Self-Distilled Intra-Relational and Progressive Information Chaining","license":"http://creativecommons.org/licenses/by/4.0/","headline":"MIPIC produces Matryoshka embeddings that remain competitive at any truncation size by aligning intra-layer relations through self-distillation and chaining semantics from deep to shallow layers.","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Hai An Vu, Linh Ngo Van, Minh-Phuc Truong, Phung Gia Huy, Thang Duc Tran, Thanh Hong Nguyen, Trung Le","submitted_at":"2026-04-27T12:07:40Z","abstract_excerpt":"Representation learning is fundamental to NLP, but building embeddings that work well at different computational budgets is challenging. Matryoshka Representation Learning (MRL) offers a flexible inference paradigm through nested embeddings; however, learning such structures requires explicit coordination of how information is arranged across embedding dimensionality and model depth. In this work, we propose MIPIC (Matryoshka Representation Learning via Self-Distilled Intra-Relational Alignment and Progressive Information Chaining), a unified training framework designed to produce structurally"},"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":"2604.24374","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-04-27T12:07:40Z","cross_cats_sorted":[],"title_canon_sha256":"c3549c1f51cffadfb142ff548c22f9d13ddb61ebe9e0459771fbfe1bc34a65a8","abstract_canon_sha256":"eda13958633f9bd9242607b6e01bc39eda462092f0b49862543e90931e316bbe"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-03T01:05:14.079431Z","signature_b64":"MXvoVw/RCfmiFpTKd872HrrrympQwPNdBVxHygmNgmEIc9D1U6UbnLyN67faaMLgY8Pew7UEroVFxY0ZydETDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"36d52be7ad61ca6cdc7faca91c809cd4cc109c3110791b51a24c53c3ce7bb0b3","last_reissued_at":"2026-06-03T01:05:14.078919Z","signature_status":"signed_v1","first_computed_at":"2026-06-03T01:05:14.078919Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MIPIC: Matryoshka Representation Learning via Self-Distilled Intra-Relational and Progressive Information Chaining","license":"http://creativecommons.org/licenses/by/4.0/","headline":"MIPIC produces Matryoshka embeddings that remain competitive at any truncation size by aligning intra-layer relations through self-distillation and chaining semantics from deep to shallow layers.","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Hai An Vu, Linh Ngo Van, Minh-Phuc Truong, Phung Gia Huy, Thang Duc Tran, Thanh Hong Nguyen, Trung Le","submitted_at":"2026-04-27T12:07:40Z","abstract_excerpt":"Representation learning is fundamental to NLP, but building embeddings that work well at different computational budgets is challenging. Matryoshka Representation Learning (MRL) offers a flexible inference paradigm through nested embeddings; however, learning such structures requires explicit coordination of how information is arranged across embedding dimensionality and model depth. In this work, we propose MIPIC (Matryoshka Representation Learning via Self-Distilled Intra-Relational Alignment and Progressive Information Chaining), a unified training framework designed to produce structurally"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"MIPIC yields Matryoshka representations that are highly competitive across all capacities, with significant performance advantages observed under extreme low-dimensional.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That aligning intra-relational structures via top-k CKA self-distillation and incrementally chaining semantics from deeper to earlier layers will produce structurally coherent and semantically compact representations without introducing distortions or requiring heavy task-specific tuning.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"MIPIC trains nested Matryoshka representations via self-distilled intra-relational alignment with top-k CKA and progressive information chaining across depths, yielding competitive performance especially at extreme low dimensions.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"MIPIC produces Matryoshka embeddings that remain competitive at any truncation size by aligning intra-layer relations through self-distillation and chaining semantics from deep to shallow layers.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"19efbc5bb6a4d17cd2131b64050eb4a67fcac1757add79f647fa2597a598f131"},"source":{"id":"2604.24374","kind":"arxiv","version":2},"verdict":{"id":"4216dd7e-6530-44dd-8fc6-4265f83ee3b5","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-08T03:31:22.754461Z","strongest_claim":"MIPIC yields Matryoshka representations that are highly competitive across all capacities, with significant performance advantages observed under extreme low-dimensional.","one_line_summary":"MIPIC trains nested Matryoshka representations via self-distilled intra-relational alignment with top-k CKA and progressive information chaining across depths, yielding competitive performance especially at extreme low dimensions.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That aligning intra-relational structures via top-k CKA self-distillation and incrementally chaining semantics from deeper to earlier layers will produce structurally coherent and semantically compact representations without introducing distortions or requiring heavy task-specific tuning.","pith_extraction_headline":"MIPIC produces Matryoshka embeddings that remain competitive at any truncation size by aligning intra-layer relations through self-distillation and chaining semantics from deep to shallow layers."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.24374/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"ai_meta_artifact","ran_at":"2026-05-21T06:41:31.633335Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T22:06:38.125027Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"34def373d91086ebef96a60c284c2ad6014e75262f99cd7a3d8030168451e316"},"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":"2604.24374","created_at":"2026-06-03T01:05:14.078987+00:00"},{"alias_kind":"arxiv_version","alias_value":"2604.24374v2","created_at":"2026-06-03T01:05:14.078987+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2604.24374","created_at":"2026-06-03T01:05:14.078987+00:00"},{"alias_kind":"pith_short_12","alias_value":"G3KSXZ5NMHFG","created_at":"2026-06-03T01:05:14.078987+00:00"},{"alias_kind":"pith_short_16","alias_value":"G3KSXZ5NMHFGZXD7","created_at":"2026-06-03T01:05:14.078987+00:00"},{"alias_kind":"pith_short_8","alias_value":"G3KSXZ5N","created_at":"2026-06-03T01:05:14.078987+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/G3KSXZ5NMHFGZXD7VSURZAE42T","json":"https://pith.science/pith/G3KSXZ5NMHFGZXD7VSURZAE42T.json","graph_json":"https://pith.science/api/pith-number/G3KSXZ5NMHFGZXD7VSURZAE42T/graph.json","events_json":"https://pith.science/api/pith-number/G3KSXZ5NMHFGZXD7VSURZAE42T/events.json","paper":"https://pith.science/paper/G3KSXZ5N"},"agent_actions":{"view_html":"https://pith.science/pith/G3KSXZ5NMHFGZXD7VSURZAE42T","download_json":"https://pith.science/pith/G3KSXZ5NMHFGZXD7VSURZAE42T.json","view_paper":"https://pith.science/paper/G3KSXZ5N","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2604.24374&json=true","fetch_graph":"https://pith.science/api/pith-number/G3KSXZ5NMHFGZXD7VSURZAE42T/graph.json","fetch_events":"https://pith.science/api/pith-number/G3KSXZ5NMHFGZXD7VSURZAE42T/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/G3KSXZ5NMHFGZXD7VSURZAE42T/action/timestamp_anchor","attest_storage":"https://pith.science/pith/G3KSXZ5NMHFGZXD7VSURZAE42T/action/storage_attestation","attest_author":"https://pith.science/pith/G3KSXZ5NMHFGZXD7VSURZAE42T/action/author_attestation","sign_citation":"https://pith.science/pith/G3KSXZ5NMHFGZXD7VSURZAE42T/action/citation_signature","submit_replication":"https://pith.science/pith/G3KSXZ5NMHFGZXD7VSURZAE42T/action/replication_record"}},"created_at":"2026-06-03T01:05:14.078987+00:00","updated_at":"2026-06-03T01:05:14.078987+00:00"}