{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:IF3DBONCMRZXTODLVSIKZGQZFK","short_pith_number":"pith:IF3DBONC","canonical_record":{"source":{"id":"2406.02528","kind":"arxiv","version":7},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-04T17:50:34Z","cross_cats_sorted":[],"title_canon_sha256":"9f8b9ce5cb0e7835779ca7f659af177224375ace2d5184683fbf6b7f9b2b9d7e","abstract_canon_sha256":"6d38102694977a457a49ced4f30842b94e40a416e9845afef57f13bc07d4abac"},"schema_version":"1.0"},"canonical_sha256":"417630b9a2647379b86bac90ac9a192ab6831a91ab1c8ee790eadbc297169369","source":{"kind":"arxiv","id":"2406.02528","version":7},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.02528","created_at":"2026-07-05T11:43:28Z"},{"alias_kind":"arxiv_version","alias_value":"2406.02528v7","created_at":"2026-07-05T11:43:28Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.02528","created_at":"2026-07-05T11:43:28Z"},{"alias_kind":"pith_short_12","alias_value":"IF3DBONCMRZX","created_at":"2026-07-05T11:43:28Z"},{"alias_kind":"pith_short_16","alias_value":"IF3DBONCMRZXTODL","created_at":"2026-07-05T11:43:28Z"},{"alias_kind":"pith_short_8","alias_value":"IF3DBONC","created_at":"2026-07-05T11:43:28Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:IF3DBONCMRZXTODLVSIKZGQZFK","target":"record","payload":{"canonical_record":{"source":{"id":"2406.02528","kind":"arxiv","version":7},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-04T17:50:34Z","cross_cats_sorted":[],"title_canon_sha256":"9f8b9ce5cb0e7835779ca7f659af177224375ace2d5184683fbf6b7f9b2b9d7e","abstract_canon_sha256":"6d38102694977a457a49ced4f30842b94e40a416e9845afef57f13bc07d4abac"},"schema_version":"1.0"},"canonical_sha256":"417630b9a2647379b86bac90ac9a192ab6831a91ab1c8ee790eadbc297169369","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:43:28.290956Z","signature_b64":"58Aunjz7WqOeb7t9xbHwaxvVXik2QAKKPyYODSJgTG/YfxW0961THNtW9gCyepC6LKDSGKqLiAT6pUS2YJ9LCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"417630b9a2647379b86bac90ac9a192ab6831a91ab1c8ee790eadbc297169369","last_reissued_at":"2026-07-05T11:43:28.290432Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:43:28.290432Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2406.02528","source_version":7,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:43:28Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"FRi127Ejp8TaAtEQUO6lxx5Y4uIPPJ4bqitHkX+o9ENUhXvsHfbCG+tvD12Dp9WdEk7YfCAODtTltKPIFboTBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-03T16:20:43.135729Z"},"content_sha256":"902f66d8fc23def1192495ace4f573388b0efbc8229f5e3aed627c513ecb5bfb","schema_version":"1.0","event_id":"sha256:902f66d8fc23def1192495ace4f573388b0efbc8229f5e3aed627c513ecb5bfb"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:IF3DBONCMRZXTODLVSIKZGQZFK","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Scalable MatMul-free Language Modeling","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Dustin Richmond, Ethan Sifferman, Jason K. Eshraghian, Peng Zhou, Rui-Jie Zhu, Steven Abreu, Sumit Bam Shrestha, Tyler Sheaves, Yiqiao Wang, Yu Zhang","submitted_at":"2024-06-04T17:50:34Z","abstract_excerpt":"Large Language Models (LLMs) have fundamentally altered how we approach scaling in machine learning. However, these models pose substantial computational and memory challenges, primarily due to the reliance on matrix multiplication (MatMul) within their attention and feed-forward (FFN) layers. We demonstrate that MatMul operations can be eliminated from LLMs while maintaining strong performance, even at billion-parameter scales. Our MatMul-free models, tested on models up to 2.7B parameters, are comparable to state-of-the-art pre-trained Transformers, and the performance gap narrows as model s"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.02528","kind":"arxiv","version":7},"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/2406.02528/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":"2026-07-05T11:43:28Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"k7HzOktLviEfOpaWIi7KiRTDKhCeAvrWxGqRld9p62qOqrskyJDfrVGiXz0OzWmikOZ8XdJGdp4JbbIhhlLtAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-03T16:20:43.136099Z"},"content_sha256":"e1165bd54fb14740543fa1ee4290839d7efa8980c2ab9fa68d25b9715d67617a","schema_version":"1.0","event_id":"sha256:e1165bd54fb14740543fa1ee4290839d7efa8980c2ab9fa68d25b9715d67617a"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/IF3DBONCMRZXTODLVSIKZGQZFK/bundle.json","state_url":"https://pith.science/pith/IF3DBONCMRZXTODLVSIKZGQZFK/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/IF3DBONCMRZXTODLVSIKZGQZFK/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":"2026-08-03T16:20:43Z","links":{"resolver":"https://pith.science/pith/IF3DBONCMRZXTODLVSIKZGQZFK","bundle":"https://pith.science/pith/IF3DBONCMRZXTODLVSIKZGQZFK/bundle.json","state":"https://pith.science/pith/IF3DBONCMRZXTODLVSIKZGQZFK/state.json","well_known_bundle":"https://pith.science/.well-known/pith/IF3DBONCMRZXTODLVSIKZGQZFK/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:IF3DBONCMRZXTODLVSIKZGQZFK","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":"6d38102694977a457a49ced4f30842b94e40a416e9845afef57f13bc07d4abac","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-04T17:50:34Z","title_canon_sha256":"9f8b9ce5cb0e7835779ca7f659af177224375ace2d5184683fbf6b7f9b2b9d7e"},"schema_version":"1.0","source":{"id":"2406.02528","kind":"arxiv","version":7}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.02528","created_at":"2026-07-05T11:43:28Z"},{"alias_kind":"arxiv_version","alias_value":"2406.02528v7","created_at":"2026-07-05T11:43:28Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.02528","created_at":"2026-07-05T11:43:28Z"},{"alias_kind":"pith_short_12","alias_value":"IF3DBONCMRZX","created_at":"2026-07-05T11:43:28Z"},{"alias_kind":"pith_short_16","alias_value":"IF3DBONCMRZXTODL","created_at":"2026-07-05T11:43:28Z"},{"alias_kind":"pith_short_8","alias_value":"IF3DBONC","created_at":"2026-07-05T11:43:28Z"}],"graph_snapshots":[{"event_id":"sha256:e1165bd54fb14740543fa1ee4290839d7efa8980c2ab9fa68d25b9715d67617a","target":"graph","created_at":"2026-07-05T11:43:28Z","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/2406.02528/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large Language Models (LLMs) have fundamentally altered how we approach scaling in machine learning. However, these models pose substantial computational and memory challenges, primarily due to the reliance on matrix multiplication (MatMul) within their attention and feed-forward (FFN) layers. We demonstrate that MatMul operations can be eliminated from LLMs while maintaining strong performance, even at billion-parameter scales. Our MatMul-free models, tested on models up to 2.7B parameters, are comparable to state-of-the-art pre-trained Transformers, and the performance gap narrows as model s","authors_text":"Dustin Richmond, Ethan Sifferman, Jason K. Eshraghian, Peng Zhou, Rui-Jie Zhu, Steven Abreu, Sumit Bam Shrestha, Tyler Sheaves, Yiqiao Wang, Yu Zhang","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-04T17:50:34Z","title":"Scalable MatMul-free Language Modeling"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.02528","kind":"arxiv","version":7},"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":"sha256:902f66d8fc23def1192495ace4f573388b0efbc8229f5e3aed627c513ecb5bfb","target":"record","created_at":"2026-07-05T11:43:28Z","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":"6d38102694977a457a49ced4f30842b94e40a416e9845afef57f13bc07d4abac","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-04T17:50:34Z","title_canon_sha256":"9f8b9ce5cb0e7835779ca7f659af177224375ace2d5184683fbf6b7f9b2b9d7e"},"schema_version":"1.0","source":{"id":"2406.02528","kind":"arxiv","version":7}},"canonical_sha256":"417630b9a2647379b86bac90ac9a192ab6831a91ab1c8ee790eadbc297169369","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"417630b9a2647379b86bac90ac9a192ab6831a91ab1c8ee790eadbc297169369","first_computed_at":"2026-07-05T11:43:28.290432Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:43:28.290432Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"58Aunjz7WqOeb7t9xbHwaxvVXik2QAKKPyYODSJgTG/YfxW0961THNtW9gCyepC6LKDSGKqLiAT6pUS2YJ9LCA==","signature_status":"signed_v1","signed_at":"2026-07-05T11:43:28.290956Z","signed_message":"canonical_sha256_bytes"},"source_id":"2406.02528","source_kind":"arxiv","source_version":7}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:902f66d8fc23def1192495ace4f573388b0efbc8229f5e3aed627c513ecb5bfb","sha256:e1165bd54fb14740543fa1ee4290839d7efa8980c2ab9fa68d25b9715d67617a"],"state_sha256":"660797b74e1dce2ca8706b7a787ec4ceee797d98a7b31eedc7e310157e9bb39f"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"MT95Rt9Ln/BBNeFmY0rTi0xmVJtHEaoJ91KR/k5GYTXsknlbf8b9cwU3Dmjy9tRCq1kxZiSZ6T4GZNwIIY5jDg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-03T16:20:43.138508Z","bundle_sha256":"6e860a1dd014b822aa0d99ccf45dba8a396604994082e050a3312f7d75ee492c"}}