{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:G7L2PM7TEHTWL2BHKMXQVD6NIM","short_pith_number":"pith:G7L2PM7T","canonical_record":{"source":{"id":"2408.15237","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-08-27T17:56:11Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"28ca1f1dd53655a90600a368bdb24ba0989f752357a41ce19f2507a29135ca99","abstract_canon_sha256":"1dab2ab50dad1e7fbe7b63f3bfeaaed867c5d3cd839ba40a3f6553ca04dd579a"},"schema_version":"1.0"},"canonical_sha256":"37d7a7b3f321e765e827532f0a8fcd43235945334e55e883b4c0ce03cc1000a3","source":{"kind":"arxiv","id":"2408.15237","version":4},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2408.15237","created_at":"2026-07-05T11:28:11Z"},{"alias_kind":"arxiv_version","alias_value":"2408.15237v4","created_at":"2026-07-05T11:28:11Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2408.15237","created_at":"2026-07-05T11:28:11Z"},{"alias_kind":"pith_short_12","alias_value":"G7L2PM7TEHTW","created_at":"2026-07-05T11:28:11Z"},{"alias_kind":"pith_short_16","alias_value":"G7L2PM7TEHTWL2BH","created_at":"2026-07-05T11:28:11Z"},{"alias_kind":"pith_short_8","alias_value":"G7L2PM7T","created_at":"2026-07-05T11:28:11Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:G7L2PM7TEHTWL2BHKMXQVD6NIM","target":"record","payload":{"canonical_record":{"source":{"id":"2408.15237","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-08-27T17:56:11Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"28ca1f1dd53655a90600a368bdb24ba0989f752357a41ce19f2507a29135ca99","abstract_canon_sha256":"1dab2ab50dad1e7fbe7b63f3bfeaaed867c5d3cd839ba40a3f6553ca04dd579a"},"schema_version":"1.0"},"canonical_sha256":"37d7a7b3f321e765e827532f0a8fcd43235945334e55e883b4c0ce03cc1000a3","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:28:11.101128Z","signature_b64":"ufaL+UFOMkFW3lC0k/mhxtmUN9iyH62w+fpJZRRHNlpZBAgS8LnbKeYE3NEE1fTiSWmbEZE5kvTroNfSVh9PDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"37d7a7b3f321e765e827532f0a8fcd43235945334e55e883b4c0ce03cc1000a3","last_reissued_at":"2026-07-05T11:28:11.100606Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:28:11.100606Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2408.15237","source_version":4,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:28:11Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"9UMUTraQdgCgWgYW50JxgTzqbglNaM8vlExBobFMmiuYaZmAlZwSvZ00zV7Kfa26MwQrQHOm8qzuIxmDBMKeDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T07:38:11.950477Z"},"content_sha256":"eb3f61b8e9d123344eba4e3be9bd0764ec6b85b88217fe3a27eb05db4e128aa2","schema_version":"1.0","event_id":"sha256:eb3f61b8e9d123344eba4e3be9bd0764ec6b85b88217fe3a27eb05db4e128aa2"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:G7L2PM7TEHTWL2BHKMXQVD6NIM","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"The Mamba in the Llama: Distilling and Accelerating Hybrid Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Alexander M. Rush, Avner May, Daniele Paliotta, Junxiong Wang, Tri Dao","submitted_at":"2024-08-27T17:56:11Z","abstract_excerpt":"Linear RNN architectures, like Mamba, can be competitive with Transformer models in language modeling while having advantageous deployment characteristics. Given the focus on training large-scale Transformer models, we consider the challenge of converting these pretrained models for deployment. We demonstrate that it is feasible to distill large Transformers into linear RNNs by reusing the linear projection weights from attention layers with academic GPU resources. The resulting hybrid model, which incorporates a quarter of the attention layers, achieves performance comparable to the original "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2408.15237","kind":"arxiv","version":4},"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/2408.15237/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":"2026-07-05T11:28:11Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"YlETD5HAiTixua7A+RFV4RzPgDPMu1kobM8wgmkKbDKIvZAI1jo7bI6aSbLepaLDO2ubVMSd90TDqas/mO+LDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T07:38:11.951497Z"},"content_sha256":"38eb9c1c0ba04f5fd06a3cd1a68bb38bf91742261cc1c04304059de122e56841","schema_version":"1.0","event_id":"sha256:38eb9c1c0ba04f5fd06a3cd1a68bb38bf91742261cc1c04304059de122e56841"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/G7L2PM7TEHTWL2BHKMXQVD6NIM/bundle.json","state_url":"https://pith.science/pith/G7L2PM7TEHTWL2BHKMXQVD6NIM/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/G7L2PM7TEHTWL2BHKMXQVD6NIM/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":"2026-08-05T07:38:11Z","links":{"resolver":"https://pith.science/pith/G7L2PM7TEHTWL2BHKMXQVD6NIM","bundle":"https://pith.science/pith/G7L2PM7TEHTWL2BHKMXQVD6NIM/bundle.json","state":"https://pith.science/pith/G7L2PM7TEHTWL2BHKMXQVD6NIM/state.json","well_known_bundle":"https://pith.science/.well-known/pith/G7L2PM7TEHTWL2BHKMXQVD6NIM/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:G7L2PM7TEHTWL2BHKMXQVD6NIM","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":"1dab2ab50dad1e7fbe7b63f3bfeaaed867c5d3cd839ba40a3f6553ca04dd579a","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-08-27T17:56:11Z","title_canon_sha256":"28ca1f1dd53655a90600a368bdb24ba0989f752357a41ce19f2507a29135ca99"},"schema_version":"1.0","source":{"id":"2408.15237","kind":"arxiv","version":4}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2408.15237","created_at":"2026-07-05T11:28:11Z"},{"alias_kind":"arxiv_version","alias_value":"2408.15237v4","created_at":"2026-07-05T11:28:11Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2408.15237","created_at":"2026-07-05T11:28:11Z"},{"alias_kind":"pith_short_12","alias_value":"G7L2PM7TEHTW","created_at":"2026-07-05T11:28:11Z"},{"alias_kind":"pith_short_16","alias_value":"G7L2PM7TEHTWL2BH","created_at":"2026-07-05T11:28:11Z"},{"alias_kind":"pith_short_8","alias_value":"G7L2PM7T","created_at":"2026-07-05T11:28:11Z"}],"graph_snapshots":[{"event_id":"sha256:38eb9c1c0ba04f5fd06a3cd1a68bb38bf91742261cc1c04304059de122e56841","target":"graph","created_at":"2026-07-05T11:28:11Z","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/2408.15237/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Linear RNN architectures, like Mamba, can be competitive with Transformer models in language modeling while having advantageous deployment characteristics. Given the focus on training large-scale Transformer models, we consider the challenge of converting these pretrained models for deployment. We demonstrate that it is feasible to distill large Transformers into linear RNNs by reusing the linear projection weights from attention layers with academic GPU resources. The resulting hybrid model, which incorporates a quarter of the attention layers, achieves performance comparable to the original ","authors_text":"Alexander M. Rush, Avner May, Daniele Paliotta, Junxiong Wang, Tri Dao","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-08-27T17:56:11Z","title":"The Mamba in the Llama: Distilling and Accelerating Hybrid Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2408.15237","kind":"arxiv","version":4},"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:eb3f61b8e9d123344eba4e3be9bd0764ec6b85b88217fe3a27eb05db4e128aa2","target":"record","created_at":"2026-07-05T11:28:11Z","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":"1dab2ab50dad1e7fbe7b63f3bfeaaed867c5d3cd839ba40a3f6553ca04dd579a","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-08-27T17:56:11Z","title_canon_sha256":"28ca1f1dd53655a90600a368bdb24ba0989f752357a41ce19f2507a29135ca99"},"schema_version":"1.0","source":{"id":"2408.15237","kind":"arxiv","version":4}},"canonical_sha256":"37d7a7b3f321e765e827532f0a8fcd43235945334e55e883b4c0ce03cc1000a3","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"37d7a7b3f321e765e827532f0a8fcd43235945334e55e883b4c0ce03cc1000a3","first_computed_at":"2026-07-05T11:28:11.100606Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:28:11.100606Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"ufaL+UFOMkFW3lC0k/mhxtmUN9iyH62w+fpJZRRHNlpZBAgS8LnbKeYE3NEE1fTiSWmbEZE5kvTroNfSVh9PDQ==","signature_status":"signed_v1","signed_at":"2026-07-05T11:28:11.101128Z","signed_message":"canonical_sha256_bytes"},"source_id":"2408.15237","source_kind":"arxiv","source_version":4}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:eb3f61b8e9d123344eba4e3be9bd0764ec6b85b88217fe3a27eb05db4e128aa2","sha256:38eb9c1c0ba04f5fd06a3cd1a68bb38bf91742261cc1c04304059de122e56841"],"state_sha256":"686a5bc07807664c668c4e5f8f7bdb5c93c40e5cf692024d8365aec6a086123b"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Mu6zRORjojTz+XCio7JXj+6cXIP2mAIs0CxxpbpRU0DLngy0Q3qPJDw8IjKzJUEDmaPQiUiRxVheVzehixrTAg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-05T07:38:11.957545Z","bundle_sha256":"a182c28a3f577179a53d825431965996fac169d09a5166d53e953542960f1045"}}