{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:DWT2XANHCKBJC72665A4EUJWDL","short_pith_number":"pith:DWT2XANH","schema_version":"1.0","canonical_sha256":"1da7ab81a71282917f5ef741c251361af80bd16acabf1112ad0d1860bd263913","source":{"kind":"arxiv","id":"2605.01374","version":2},"attestation_state":"computed","paper":{"title":"MTA: Multi-Granular Trajectory Alignment for Large Language Model Distillation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Multi-granular trajectory alignment improves knowledge distillation by matching teacher and student representations at word level in lower layers and phrase level in higher layers.","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Linh Ngo Van, Pham Khanh Chi, Quoc Phong Dao, Thanh Hong Nguyen, Thuat Nguyen, Trung Le","submitted_at":"2026-05-02T10:52:49Z","abstract_excerpt":"Knowledge distillation is a key technique for compressing large language models (LLMs), but most existing methods align representations at fixed layers or token-level outputs, ignoring how representations evolve across depth. As a result, the student is only weakly guided to capture the teacher's internal relational structure during distillation, which limits knowledge transfer. To address this limitation, we propose Multi-Granular Trajectory Alignment (MTA), a framework that aligns teacher and student representations along their layer-wise transformation trajectory. MTA adopts a layer-adaptiv"},"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":"2605.01374","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-02T10:52:49Z","cross_cats_sorted":[],"title_canon_sha256":"1b270c6c37c74de47904b5efb3f575514dfda75bb956d0009de8d7354a426b05","abstract_canon_sha256":"4a209011ec524ab89114606c2795774c10a891a1e4d1f64f66f42de78d4ce61e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-03T01:05:50.831123Z","signature_b64":"RMG0FJ86eN4K5ejLQWhC5FSyy1POhgeL/j8IPUJlWn1OldNOBvhtXoEmIA5+rNoaAeI/6mJRXWduNWqotM75Aw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1da7ab81a71282917f5ef741c251361af80bd16acabf1112ad0d1860bd263913","last_reissued_at":"2026-06-03T01:05:50.830716Z","signature_status":"signed_v1","first_computed_at":"2026-06-03T01:05:50.830716Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MTA: Multi-Granular Trajectory Alignment for Large Language Model Distillation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Multi-granular trajectory alignment improves knowledge distillation by matching teacher and student representations at word level in lower layers and phrase level in higher layers.","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Linh Ngo Van, Pham Khanh Chi, Quoc Phong Dao, Thanh Hong Nguyen, Thuat Nguyen, Trung Le","submitted_at":"2026-05-02T10:52:49Z","abstract_excerpt":"Knowledge distillation is a key technique for compressing large language models (LLMs), but most existing methods align representations at fixed layers or token-level outputs, ignoring how representations evolve across depth. As a result, the student is only weakly guided to capture the teacher's internal relational structure during distillation, which limits knowledge transfer. To address this limitation, we propose Multi-Granular Trajectory Alignment (MTA), a framework that aligns teacher and student representations along their layer-wise transformation trajectory. MTA adopts a layer-adaptiv"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"MTA consistently outperforms state-of-the-art baselines on standard benchmarks, with ablations confirming the contribution of each component.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That aligning teacher and student representations along their layer-wise transformation trajectory using a layer-adaptive multi-granular strategy (word-level lower, phrase-level higher) will better guide the student to capture the teacher's internal relational structure than fixed-layer or token-level methods.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"MTA improves LLM knowledge distillation by aligning representations along layer-wise trajectories with adaptive granularity from words to phrases using dynamic structural and hidden representation alignment losses.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Multi-granular trajectory alignment improves knowledge distillation by matching teacher and student representations at word level in lower layers and phrase level in higher layers.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"054a12ff34e7f9874c0f1e6053c3c6a0dbd8e11508c2f494a714da33f122e632"},"source":{"id":"2605.01374","kind":"arxiv","version":2},"verdict":{"id":"0aa38e82-e6fa-4fdd-9d6e-d325616a4875","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-09T14:50:59.346561Z","strongest_claim":"MTA consistently outperforms state-of-the-art baselines on standard benchmarks, with ablations confirming the contribution of each component.","one_line_summary":"MTA improves LLM knowledge distillation by aligning representations along layer-wise trajectories with adaptive granularity from words to phrases using dynamic structural and hidden representation alignment losses.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That aligning teacher and student representations along their layer-wise transformation trajectory using a layer-adaptive multi-granular strategy (word-level lower, phrase-level higher) will better guide the student to capture the teacher's internal relational structure than fixed-layer or token-level methods.","pith_extraction_headline":"Multi-granular trajectory alignment improves knowledge distillation by matching teacher and student representations at word level in lower layers and phrase level in higher layers."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.01374/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"ai_meta_artifact","ran_at":"2026-05-20T18:34:27.146335Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T17:19:42.068043Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"9ae7dd55168262a410c47f3b61079f92c9d350a401bbe530ed675e3582b7d7aa"},"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":"2605.01374","created_at":"2026-06-03T01:05:50.830770+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.01374v2","created_at":"2026-06-03T01:05:50.830770+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.01374","created_at":"2026-06-03T01:05:50.830770+00:00"},{"alias_kind":"pith_short_12","alias_value":"DWT2XANHCKBJ","created_at":"2026-06-03T01:05:50.830770+00:00"},{"alias_kind":"pith_short_16","alias_value":"DWT2XANHCKBJC726","created_at":"2026-06-03T01:05:50.830770+00:00"},{"alias_kind":"pith_short_8","alias_value":"DWT2XANH","created_at":"2026-06-03T01:05:50.830770+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/DWT2XANHCKBJC72665A4EUJWDL","json":"https://pith.science/pith/DWT2XANHCKBJC72665A4EUJWDL.json","graph_json":"https://pith.science/api/pith-number/DWT2XANHCKBJC72665A4EUJWDL/graph.json","events_json":"https://pith.science/api/pith-number/DWT2XANHCKBJC72665A4EUJWDL/events.json","paper":"https://pith.science/paper/DWT2XANH"},"agent_actions":{"view_html":"https://pith.science/pith/DWT2XANHCKBJC72665A4EUJWDL","download_json":"https://pith.science/pith/DWT2XANHCKBJC72665A4EUJWDL.json","view_paper":"https://pith.science/paper/DWT2XANH","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.01374&json=true","fetch_graph":"https://pith.science/api/pith-number/DWT2XANHCKBJC72665A4EUJWDL/graph.json","fetch_events":"https://pith.science/api/pith-number/DWT2XANHCKBJC72665A4EUJWDL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DWT2XANHCKBJC72665A4EUJWDL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DWT2XANHCKBJC72665A4EUJWDL/action/storage_attestation","attest_author":"https://pith.science/pith/DWT2XANHCKBJC72665A4EUJWDL/action/author_attestation","sign_citation":"https://pith.science/pith/DWT2XANHCKBJC72665A4EUJWDL/action/citation_signature","submit_replication":"https://pith.science/pith/DWT2XANHCKBJC72665A4EUJWDL/action/replication_record"}},"created_at":"2026-06-03T01:05:50.830770+00:00","updated_at":"2026-06-03T01:05:50.830770+00:00"}