{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:ZCYPOSVLFNEO4E47MJP7SPF2Z3","short_pith_number":"pith:ZCYPOSVL","canonical_record":{"source":{"id":"2309.08708","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-09-15T19:00:00Z","cross_cats_sorted":[],"title_canon_sha256":"e75f32beed959d490c0da83975edd10103e76d15fb79160e0f8e0616871a8843","abstract_canon_sha256":"d824031d85f3dfeac3db1572ab5f2be976a5b12da0b3a783fc4cd1bf36049ef8"},"schema_version":"1.0"},"canonical_sha256":"c8b0f74aab2b48ee139f625ff93cbacec84038333b5a2cb2d017eaa614b2e5cf","source":{"kind":"arxiv","id":"2309.08708","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2309.08708","created_at":"2026-07-05T10:38:31Z"},{"alias_kind":"arxiv_version","alias_value":"2309.08708v2","created_at":"2026-07-05T10:38:31Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2309.08708","created_at":"2026-07-05T10:38:31Z"},{"alias_kind":"pith_short_12","alias_value":"ZCYPOSVLFNEO","created_at":"2026-07-05T10:38:31Z"},{"alias_kind":"pith_short_16","alias_value":"ZCYPOSVLFNEO4E47","created_at":"2026-07-05T10:38:31Z"},{"alias_kind":"pith_short_8","alias_value":"ZCYPOSVL","created_at":"2026-07-05T10:38:31Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:ZCYPOSVLFNEO4E47MJP7SPF2Z3","target":"record","payload":{"canonical_record":{"source":{"id":"2309.08708","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-09-15T19:00:00Z","cross_cats_sorted":[],"title_canon_sha256":"e75f32beed959d490c0da83975edd10103e76d15fb79160e0f8e0616871a8843","abstract_canon_sha256":"d824031d85f3dfeac3db1572ab5f2be976a5b12da0b3a783fc4cd1bf36049ef8"},"schema_version":"1.0"},"canonical_sha256":"c8b0f74aab2b48ee139f625ff93cbacec84038333b5a2cb2d017eaa614b2e5cf","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:38:31.106759Z","signature_b64":"yrGgoyVtU1fxmNkh7pVvtv7Oy/yVygGjsvzNKhJF6wkqduDGtZY2Jo50Sgg3KvPOgXyltkKYY9TzaIMw1KtmDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c8b0f74aab2b48ee139f625ff93cbacec84038333b5a2cb2d017eaa614b2e5cf","last_reissued_at":"2026-07-05T10:38:31.106347Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:38:31.106347Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2309.08708","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:38:31Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Do3v6oRvKjszpln+aGSyaOe6op4C129A78Lv7FkuKvQl7TJIUt0R4H7LAfHLworfy13cB8rCgy6ZRXZHwJHyDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T06:54:40.199356Z"},"content_sha256":"a291de28f2f44b1ec8784903ffbf00bc473fb189123303fd51c7c1bd47297c6d","schema_version":"1.0","event_id":"sha256:a291de28f2f44b1ec8784903ffbf00bc473fb189123303fd51c7c1bd47297c6d"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:ZCYPOSVLFNEO4E47MJP7SPF2Z3","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Vocabulary-level Memory Efficiency for Language Model Fine-tuning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Miles Williams, Nikolaos Aletras","submitted_at":"2023-09-15T19:00:00Z","abstract_excerpt":"The extensive memory footprint of language model (LM) fine-tuning poses a challenge for both researchers and practitioners. LMs use an embedding matrix to represent extensive vocabularies, forming a substantial proportion of the model parameters. While previous work towards memory-efficient fine-tuning has focused on minimizing the number of trainable parameters, reducing the memory footprint of the embedding matrix has yet to be explored. We first demonstrate that a significant proportion of the vocabulary remains unused during fine-tuning. We then propose a simple yet effective approach that"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2309.08708","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/2309.08708/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":"2026-07-05T10:38:31Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"p+ggMkai6Q/1I5zYs3eCokGdacGmQ4dPbgWrthV4mTGAHNgqpFz3QySn7EpkZOmk0dhXXCEFngoR8W4tJJNaAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T06:54:40.200263Z"},"content_sha256":"db7e398f6d1f33dc3cb441843d0cb5ee82b7db9ec717907c82956260bc0bea38","schema_version":"1.0","event_id":"sha256:db7e398f6d1f33dc3cb441843d0cb5ee82b7db9ec717907c82956260bc0bea38"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/ZCYPOSVLFNEO4E47MJP7SPF2Z3/bundle.json","state_url":"https://pith.science/pith/ZCYPOSVLFNEO4E47MJP7SPF2Z3/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/ZCYPOSVLFNEO4E47MJP7SPF2Z3/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":"2026-08-08T06:54:40Z","links":{"resolver":"https://pith.science/pith/ZCYPOSVLFNEO4E47MJP7SPF2Z3","bundle":"https://pith.science/pith/ZCYPOSVLFNEO4E47MJP7SPF2Z3/bundle.json","state":"https://pith.science/pith/ZCYPOSVLFNEO4E47MJP7SPF2Z3/state.json","well_known_bundle":"https://pith.science/.well-known/pith/ZCYPOSVLFNEO4E47MJP7SPF2Z3/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:ZCYPOSVLFNEO4E47MJP7SPF2Z3","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":"d824031d85f3dfeac3db1572ab5f2be976a5b12da0b3a783fc4cd1bf36049ef8","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-09-15T19:00:00Z","title_canon_sha256":"e75f32beed959d490c0da83975edd10103e76d15fb79160e0f8e0616871a8843"},"schema_version":"1.0","source":{"id":"2309.08708","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2309.08708","created_at":"2026-07-05T10:38:31Z"},{"alias_kind":"arxiv_version","alias_value":"2309.08708v2","created_at":"2026-07-05T10:38:31Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2309.08708","created_at":"2026-07-05T10:38:31Z"},{"alias_kind":"pith_short_12","alias_value":"ZCYPOSVLFNEO","created_at":"2026-07-05T10:38:31Z"},{"alias_kind":"pith_short_16","alias_value":"ZCYPOSVLFNEO4E47","created_at":"2026-07-05T10:38:31Z"},{"alias_kind":"pith_short_8","alias_value":"ZCYPOSVL","created_at":"2026-07-05T10:38:31Z"}],"graph_snapshots":[{"event_id":"sha256:db7e398f6d1f33dc3cb441843d0cb5ee82b7db9ec717907c82956260bc0bea38","target":"graph","created_at":"2026-07-05T10:38:31Z","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/2309.08708/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The extensive memory footprint of language model (LM) fine-tuning poses a challenge for both researchers and practitioners. LMs use an embedding matrix to represent extensive vocabularies, forming a substantial proportion of the model parameters. While previous work towards memory-efficient fine-tuning has focused on minimizing the number of trainable parameters, reducing the memory footprint of the embedding matrix has yet to be explored. We first demonstrate that a significant proportion of the vocabulary remains unused during fine-tuning. We then propose a simple yet effective approach that","authors_text":"Miles Williams, Nikolaos Aletras","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-09-15T19:00:00Z","title":"Vocabulary-level Memory Efficiency for Language Model Fine-tuning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2309.08708","kind":"arxiv","version":2},"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:a291de28f2f44b1ec8784903ffbf00bc473fb189123303fd51c7c1bd47297c6d","target":"record","created_at":"2026-07-05T10:38:31Z","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":"d824031d85f3dfeac3db1572ab5f2be976a5b12da0b3a783fc4cd1bf36049ef8","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-09-15T19:00:00Z","title_canon_sha256":"e75f32beed959d490c0da83975edd10103e76d15fb79160e0f8e0616871a8843"},"schema_version":"1.0","source":{"id":"2309.08708","kind":"arxiv","version":2}},"canonical_sha256":"c8b0f74aab2b48ee139f625ff93cbacec84038333b5a2cb2d017eaa614b2e5cf","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"c8b0f74aab2b48ee139f625ff93cbacec84038333b5a2cb2d017eaa614b2e5cf","first_computed_at":"2026-07-05T10:38:31.106347Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:38:31.106347Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"yrGgoyVtU1fxmNkh7pVvtv7Oy/yVygGjsvzNKhJF6wkqduDGtZY2Jo50Sgg3KvPOgXyltkKYY9TzaIMw1KtmDA==","signature_status":"signed_v1","signed_at":"2026-07-05T10:38:31.106759Z","signed_message":"canonical_sha256_bytes"},"source_id":"2309.08708","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:a291de28f2f44b1ec8784903ffbf00bc473fb189123303fd51c7c1bd47297c6d","sha256:db7e398f6d1f33dc3cb441843d0cb5ee82b7db9ec717907c82956260bc0bea38"],"state_sha256":"d5ca59b7d95457ce97ce37cd2fec37db43f4329d23504612802bf5bbcd701fe3"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"o3wFQKqwlCdH1585Gg1kzAWNnqKt2uiFffXQ3RhKxdtldFrYFvC+7PPoWl5MkcB4OmaxnbaLdg+G26B6Lz5xAg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-08T06:54:40.205386Z","bundle_sha256":"629277e61ed5cd14326f98b712dec66ef03fbb5ee01706e24de495399abb14e1"}}