{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:DAQZWHZLK4F3TZNWBPO7UCREOQ","short_pith_number":"pith:DAQZWHZL","schema_version":"1.0","canonical_sha256":"18219b1f2b570bb9e5b60bddfa0a24741ab0ea339cff35547da0c65da2d2ce2e","source":{"kind":"arxiv","id":"2408.10189","version":2},"attestation_state":"computed","paper":{"title":"Transformers to SSMs: Distilling Quadratic Knowledge to Subquadratic Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Albert Gu, Aviv Bick, Eric P. Xing, J. Zico Kolter, Kevin Y. Li","submitted_at":"2024-08-19T17:48:11Z","abstract_excerpt":"Transformer architectures have become a dominant paradigm for domains like language modeling but suffer in many inference settings due to their quadratic-time self-attention. Recently proposed subquadratic architectures, such as Mamba, have shown promise, but have been pretrained with substantially less computational resources than the strongest Transformer models. In this work, we present a method that is able to distill a pretrained Transformer architecture into alternative architectures such as state space models (SSMs). The key idea to our approach is that we can view both Transformers and"},"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":"2408.10189","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-08-19T17:48:11Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"6426238307fdd96fa0a76d17e628ed893ab4de964f677d4b09da4cc4f3e2d677","abstract_canon_sha256":"f7fa038d62418447d24285240533171fd0b73b37ad80233f277384a9cd17020f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:11:23.194706Z","signature_b64":"cZMm6okBjblu5XSGhVdm0QwZZwRm7LeJt5BxaFrPF9jZ1qjQlcTtVm2TJ1vaa+hVuZka3kt0w9cHNPmhRt43Bw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"18219b1f2b570bb9e5b60bddfa0a24741ab0ea339cff35547da0c65da2d2ce2e","last_reissued_at":"2026-07-05T10:11:23.194199Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:11:23.194199Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Transformers to SSMs: Distilling Quadratic Knowledge to Subquadratic Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Albert Gu, Aviv Bick, Eric P. Xing, J. Zico Kolter, Kevin Y. Li","submitted_at":"2024-08-19T17:48:11Z","abstract_excerpt":"Transformer architectures have become a dominant paradigm for domains like language modeling but suffer in many inference settings due to their quadratic-time self-attention. Recently proposed subquadratic architectures, such as Mamba, have shown promise, but have been pretrained with substantially less computational resources than the strongest Transformer models. In this work, we present a method that is able to distill a pretrained Transformer architecture into alternative architectures such as state space models (SSMs). The key idea to our approach is that we can view both Transformers and"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2408.10189","kind":"arxiv","version":2},"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/2408.10189/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":"2408.10189","created_at":"2026-07-05T10:11:23.194257+00:00"},{"alias_kind":"arxiv_version","alias_value":"2408.10189v2","created_at":"2026-07-05T10:11:23.194257+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2408.10189","created_at":"2026-07-05T10:11:23.194257+00:00"},{"alias_kind":"pith_short_12","alias_value":"DAQZWHZLK4F3","created_at":"2026-07-05T10:11:23.194257+00:00"},{"alias_kind":"pith_short_16","alias_value":"DAQZWHZLK4F3TZNW","created_at":"2026-07-05T10:11:23.194257+00:00"},{"alias_kind":"pith_short_8","alias_value":"DAQZWHZL","created_at":"2026-07-05T10:11:23.194257+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":8,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.19867","citing_title":"PSCT-Net: Geometry-Aware Pediatric Skull CT Reconstruction via Differentiable Back-Projection and Attention-Guided Refinement","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2606.19867","citing_title":"PSCT-Net: Geometry-Aware Pediatric Skull CT Reconstruction via Differentiable Back-Projection and Attention-Guided Refinement","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2606.07604","citing_title":"Contribution Weights: A Geometrical Analysis of Self-Attention Transformers","ref_index":128,"is_internal_anchor":false},{"citing_arxiv_id":"2606.19867","citing_title":"PSCT-Net: Geometry-Aware Pediatric Skull CT Reconstruction via Differentiable Back-Projection and Attention-Guided Refinement","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2410.13846","citing_title":"LightTransfer: Your Long-Context LLM is Secretly a Hybrid Model with Effortless Adaptation","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2509.04154","citing_title":"Robust Filter Attention: Self-Attention as Precision-Weighted State Estimation","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11007","citing_title":"The Transformer as a Polar State Estimator","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2604.24715","citing_title":"Long-Context Aware Upcycling: A New Frontier for Hybrid LLM Scaling","ref_index":1,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DAQZWHZLK4F3TZNWBPO7UCREOQ","json":"https://pith.science/pith/DAQZWHZLK4F3TZNWBPO7UCREOQ.json","graph_json":"https://pith.science/api/pith-number/DAQZWHZLK4F3TZNWBPO7UCREOQ/graph.json","events_json":"https://pith.science/api/pith-number/DAQZWHZLK4F3TZNWBPO7UCREOQ/events.json","paper":"https://pith.science/paper/DAQZWHZL"},"agent_actions":{"view_html":"https://pith.science/pith/DAQZWHZLK4F3TZNWBPO7UCREOQ","download_json":"https://pith.science/pith/DAQZWHZLK4F3TZNWBPO7UCREOQ.json","view_paper":"https://pith.science/paper/DAQZWHZL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2408.10189&json=true","fetch_graph":"https://pith.science/api/pith-number/DAQZWHZLK4F3TZNWBPO7UCREOQ/graph.json","fetch_events":"https://pith.science/api/pith-number/DAQZWHZLK4F3TZNWBPO7UCREOQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DAQZWHZLK4F3TZNWBPO7UCREOQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DAQZWHZLK4F3TZNWBPO7UCREOQ/action/storage_attestation","attest_author":"https://pith.science/pith/DAQZWHZLK4F3TZNWBPO7UCREOQ/action/author_attestation","sign_citation":"https://pith.science/pith/DAQZWHZLK4F3TZNWBPO7UCREOQ/action/citation_signature","submit_replication":"https://pith.science/pith/DAQZWHZLK4F3TZNWBPO7UCREOQ/action/replication_record"}},"created_at":"2026-07-05T10:11:23.194257+00:00","updated_at":"2026-07-05T10:11:23.194257+00:00"}