{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:6MJZWBG3TPKNQPIZMO43CIV7UY","short_pith_number":"pith:6MJZWBG3","canonical_record":{"source":{"id":"2410.10254","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-10-14T08:10:34Z","cross_cats_sorted":["cs.AI","cs.CL","stat.ML"],"title_canon_sha256":"1429d0a450a16eea11cdb0d596c2b66392b99eb570263bf2f89514cd5d54fe37","abstract_canon_sha256":"75cafaadf3adaaaa306139a842bd32ceffade60cc47e668f0b4c7baffde8ca1d"},"schema_version":"1.0"},"canonical_sha256":"f3139b04db9bd4d83d1963b9b122bfa601134c53f991aa039c788700a48fd215","source":{"kind":"arxiv","id":"2410.10254","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.10254","created_at":"2026-07-05T10:25:11Z"},{"alias_kind":"arxiv_version","alias_value":"2410.10254v3","created_at":"2026-07-05T10:25:11Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.10254","created_at":"2026-07-05T10:25:11Z"},{"alias_kind":"pith_short_12","alias_value":"6MJZWBG3TPKN","created_at":"2026-07-05T10:25:11Z"},{"alias_kind":"pith_short_16","alias_value":"6MJZWBG3TPKNQPIZ","created_at":"2026-07-05T10:25:11Z"},{"alias_kind":"pith_short_8","alias_value":"6MJZWBG3","created_at":"2026-07-05T10:25:11Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:6MJZWBG3TPKNQPIZMO43CIV7UY","target":"record","payload":{"canonical_record":{"source":{"id":"2410.10254","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-10-14T08:10:34Z","cross_cats_sorted":["cs.AI","cs.CL","stat.ML"],"title_canon_sha256":"1429d0a450a16eea11cdb0d596c2b66392b99eb570263bf2f89514cd5d54fe37","abstract_canon_sha256":"75cafaadf3adaaaa306139a842bd32ceffade60cc47e668f0b4c7baffde8ca1d"},"schema_version":"1.0"},"canonical_sha256":"f3139b04db9bd4d83d1963b9b122bfa601134c53f991aa039c788700a48fd215","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:25:11.673204Z","signature_b64":"tAFfcQEL/sYSxaV7yimg1+AEyjkdzgwu4/YLT8yCmihDL9ysP1wEo0qwG2wvUVQ2Os4ZGsbmC7cf8HZo7de+Cg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f3139b04db9bd4d83d1963b9b122bfa601134c53f991aa039c788700a48fd215","last_reissued_at":"2026-07-05T10:25:11.672126Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:25:11.672126Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2410.10254","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:25:11Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"iKPI4KqYsg4WcN0gJemKqmoYEwypWjvsZgoEjMHVPnztDpA5Ww1xg/5ZdpAghGJSvOmrI8DlLrXBWaFXD1ENBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T02:45:27.636912Z"},"content_sha256":"7401dc63808f4fe33c6c6e07aae3356a7278dae5fdd3245a8f17ce014e83a7a1","schema_version":"1.0","event_id":"sha256:7401dc63808f4fe33c6c6e07aae3356a7278dae5fdd3245a8f17ce014e83a7a1"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:6MJZWBG3TPKNQPIZMO43CIV7UY","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"LoLCATs: On Low-Rank Linearizing of Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL","stat.ML"],"primary_cat":"cs.LG","authors_text":"Aaryan Singhal, Alan Wu, Benjamin Spector, Christopher R\\'e, Krithik Ramesh, Michael Zhang, Rahul Chalamala, Simran Arora","submitted_at":"2024-10-14T08:10:34Z","abstract_excerpt":"Recent works show we can linearize large language models (LLMs) -- swapping the quadratic attentions of popular Transformer-based LLMs with subquadratic analogs, such as linear attention -- avoiding the expensive pretraining costs. However, linearizing LLMs often significantly degrades model quality, still requires training over billions of tokens, and remains limited to smaller 1.3B to 7B LLMs. We thus propose Low-rank Linear Conversion via Attention Transfer (LoLCATs), a simple two-step method that improves LLM linearizing quality with orders of magnitudes less memory and compute. We base th"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.10254","kind":"arxiv","version":3},"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/2410.10254/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:25:11Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"up7+Du5+i2DEP1uyJ4gmWsoFiSTaf4e9vnP7Zd6I8OcXKjUgtztP+Kpr0Jx2nVaz68jnCSlx31UwkGePpfU2Cw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T02:45:27.637433Z"},"content_sha256":"0c9ff05cf178bd47b11225b224903b7edcf7ccc7067b028599aa5b37e8940541","schema_version":"1.0","event_id":"sha256:0c9ff05cf178bd47b11225b224903b7edcf7ccc7067b028599aa5b37e8940541"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/6MJZWBG3TPKNQPIZMO43CIV7UY/bundle.json","state_url":"https://pith.science/pith/6MJZWBG3TPKNQPIZMO43CIV7UY/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/6MJZWBG3TPKNQPIZMO43CIV7UY/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-06T02:45:27Z","links":{"resolver":"https://pith.science/pith/6MJZWBG3TPKNQPIZMO43CIV7UY","bundle":"https://pith.science/pith/6MJZWBG3TPKNQPIZMO43CIV7UY/bundle.json","state":"https://pith.science/pith/6MJZWBG3TPKNQPIZMO43CIV7UY/state.json","well_known_bundle":"https://pith.science/.well-known/pith/6MJZWBG3TPKNQPIZMO43CIV7UY/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:6MJZWBG3TPKNQPIZMO43CIV7UY","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"75cafaadf3adaaaa306139a842bd32ceffade60cc47e668f0b4c7baffde8ca1d","cross_cats_sorted":["cs.AI","cs.CL","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-10-14T08:10:34Z","title_canon_sha256":"1429d0a450a16eea11cdb0d596c2b66392b99eb570263bf2f89514cd5d54fe37"},"schema_version":"1.0","source":{"id":"2410.10254","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.10254","created_at":"2026-07-05T10:25:11Z"},{"alias_kind":"arxiv_version","alias_value":"2410.10254v3","created_at":"2026-07-05T10:25:11Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.10254","created_at":"2026-07-05T10:25:11Z"},{"alias_kind":"pith_short_12","alias_value":"6MJZWBG3TPKN","created_at":"2026-07-05T10:25:11Z"},{"alias_kind":"pith_short_16","alias_value":"6MJZWBG3TPKNQPIZ","created_at":"2026-07-05T10:25:11Z"},{"alias_kind":"pith_short_8","alias_value":"6MJZWBG3","created_at":"2026-07-05T10:25:11Z"}],"graph_snapshots":[{"event_id":"sha256:0c9ff05cf178bd47b11225b224903b7edcf7ccc7067b028599aa5b37e8940541","target":"graph","created_at":"2026-07-05T10:25:11Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2410.10254/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Recent works show we can linearize large language models (LLMs) -- swapping the quadratic attentions of popular Transformer-based LLMs with subquadratic analogs, such as linear attention -- avoiding the expensive pretraining costs. However, linearizing LLMs often significantly degrades model quality, still requires training over billions of tokens, and remains limited to smaller 1.3B to 7B LLMs. We thus propose Low-rank Linear Conversion via Attention Transfer (LoLCATs), a simple two-step method that improves LLM linearizing quality with orders of magnitudes less memory and compute. We base th","authors_text":"Aaryan Singhal, Alan Wu, Benjamin Spector, Christopher R\\'e, Krithik Ramesh, Michael Zhang, Rahul Chalamala, Simran Arora","cross_cats":["cs.AI","cs.CL","stat.ML"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-10-14T08:10:34Z","title":"LoLCATs: On Low-Rank Linearizing of Large Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.10254","kind":"arxiv","version":3},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:7401dc63808f4fe33c6c6e07aae3356a7278dae5fdd3245a8f17ce014e83a7a1","target":"record","created_at":"2026-07-05T10:25:11Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"75cafaadf3adaaaa306139a842bd32ceffade60cc47e668f0b4c7baffde8ca1d","cross_cats_sorted":["cs.AI","cs.CL","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-10-14T08:10:34Z","title_canon_sha256":"1429d0a450a16eea11cdb0d596c2b66392b99eb570263bf2f89514cd5d54fe37"},"schema_version":"1.0","source":{"id":"2410.10254","kind":"arxiv","version":3}},"canonical_sha256":"f3139b04db9bd4d83d1963b9b122bfa601134c53f991aa039c788700a48fd215","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f3139b04db9bd4d83d1963b9b122bfa601134c53f991aa039c788700a48fd215","first_computed_at":"2026-07-05T10:25:11.672126Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:25:11.672126Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"tAFfcQEL/sYSxaV7yimg1+AEyjkdzgwu4/YLT8yCmihDL9ysP1wEo0qwG2wvUVQ2Os4ZGsbmC7cf8HZo7de+Cg==","signature_status":"signed_v1","signed_at":"2026-07-05T10:25:11.673204Z","signed_message":"canonical_sha256_bytes"},"source_id":"2410.10254","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:7401dc63808f4fe33c6c6e07aae3356a7278dae5fdd3245a8f17ce014e83a7a1","sha256:0c9ff05cf178bd47b11225b224903b7edcf7ccc7067b028599aa5b37e8940541"],"state_sha256":"01f165485898c5f580622706653a3f3b01cabd2c6003e807c02d3d3c0e364945"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"0JPHFuoe3RG6unDbhy0Tns5U2jXDT9WUNfxU/Kn5AOaNBSYGBJgikxzwfR99HXkUVCajmS8v1rmh6yceVBqeDw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-06T02:45:27.641068Z","bundle_sha256":"925c40c59395ad9a470c9a2ea02e596586e00e0ad5591bb252de5df27179163f"}}