{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:QTJGAUGHWPJF7TS6AO4IIBA3RP","short_pith_number":"pith:QTJGAUGH","canonical_record":{"source":{"id":"2505.06150","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-09T16:02:23Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"05ac7fb5cd4da151cfb073a70370c9c2d698939591d5493cd8f040819911cc64","abstract_canon_sha256":"9bd95205037c05b650e01702bd7142f2c651cd1f7ca8dbe9aac2b602f5e56778"},"schema_version":"1.0"},"canonical_sha256":"84d26050c7b3d25fce5e03b884041b8bd814e1fe582a64501429c030261335ca","source":{"kind":"arxiv","id":"2505.06150","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.06150","created_at":"2026-07-05T11:14:28Z"},{"alias_kind":"arxiv_version","alias_value":"2505.06150v2","created_at":"2026-07-05T11:14:28Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.06150","created_at":"2026-07-05T11:14:28Z"},{"alias_kind":"pith_short_12","alias_value":"QTJGAUGHWPJF","created_at":"2026-07-05T11:14:28Z"},{"alias_kind":"pith_short_16","alias_value":"QTJGAUGHWPJF7TS6","created_at":"2026-07-05T11:14:28Z"},{"alias_kind":"pith_short_8","alias_value":"QTJGAUGH","created_at":"2026-07-05T11:14:28Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:QTJGAUGHWPJF7TS6AO4IIBA3RP","target":"record","payload":{"canonical_record":{"source":{"id":"2505.06150","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-09T16:02:23Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"05ac7fb5cd4da151cfb073a70370c9c2d698939591d5493cd8f040819911cc64","abstract_canon_sha256":"9bd95205037c05b650e01702bd7142f2c651cd1f7ca8dbe9aac2b602f5e56778"},"schema_version":"1.0"},"canonical_sha256":"84d26050c7b3d25fce5e03b884041b8bd814e1fe582a64501429c030261335ca","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:14:28.919851Z","signature_b64":"Z26b7CGb7eiU3/DNIK2Dx2nfwESEgs7f9uFcdIFdBUEuesQkS06owWj6GsU4vGszjENgi/uYm4fm50gLUsOjDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"84d26050c7b3d25fce5e03b884041b8bd814e1fe582a64501429c030261335ca","last_reissued_at":"2026-07-05T11:14:28.919373Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:14:28.919373Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2505.06150","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":"2026-07-05T11:14:28Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"G07JTuoapio71Gtl20Wsz/l+dLIBXvnw+F+48rHxwrkfEeiTnUqxKOstoXtt7134P/flhiIHeX4MCwKfVT13Bg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-23T22:00:31.466456Z"},"content_sha256":"5fa46d06f7d0cb38c5d688bbb84ce5fe26b435d0627029a008faf00639415e13","schema_version":"1.0","event_id":"sha256:5fa46d06f7d0cb38c5d688bbb84ce5fe26b435d0627029a008faf00639415e13"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:QTJGAUGHWPJF7TS6AO4IIBA3RP","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"A Scaling Law for Token Efficiency in LLM Fine-Tuning Under Fixed Compute Budgets","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Aidan Kierans, Avijit Ghosh, Ryan Lagasse, Shiri Dori-Hacohen","submitted_at":"2025-05-09T16:02:23Z","abstract_excerpt":"We introduce a scaling law for fine-tuning large language models (LLMs) under fixed compute budgets that explicitly accounts for data composition. Conventional approaches measure training data solely by total tokens, yet the number of examples and their average token length -- what we term \\emph{dataset volume} -- play a decisive role in model performance. Our formulation is tuned following established procedures. Experiments on the BRICC dataset \\cite{salavati2024reducing} and subsets of the MMLU dataset \\cite{hendrycks2021measuringmassivemultitasklanguage}, evaluated under multiple subsampli"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.06150","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/2505.06150/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":"2026-07-05T11:14:28Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"SCu9IWcQHzSn1GnhbhwtqBT748gEw9MDlwKWiwzfxFt4FpVNyXlAGRBHJLIs0RAwr97qL+mDHtifhQfo42ffBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-23T22:00:31.466957Z"},"content_sha256":"dce11f14ddbe18fff1a54e04ee204d1d09fd4191b6d805d2e139e22ffc4cfcfa","schema_version":"1.0","event_id":"sha256:dce11f14ddbe18fff1a54e04ee204d1d09fd4191b6d805d2e139e22ffc4cfcfa"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/QTJGAUGHWPJF7TS6AO4IIBA3RP/bundle.json","state_url":"https://pith.science/pith/QTJGAUGHWPJF7TS6AO4IIBA3RP/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/QTJGAUGHWPJF7TS6AO4IIBA3RP/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":"2026-08-23T22:00:31Z","links":{"resolver":"https://pith.science/pith/QTJGAUGHWPJF7TS6AO4IIBA3RP","bundle":"https://pith.science/pith/QTJGAUGHWPJF7TS6AO4IIBA3RP/bundle.json","state":"https://pith.science/pith/QTJGAUGHWPJF7TS6AO4IIBA3RP/state.json","well_known_bundle":"https://pith.science/.well-known/pith/QTJGAUGHWPJF7TS6AO4IIBA3RP/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:QTJGAUGHWPJF7TS6AO4IIBA3RP","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":"9bd95205037c05b650e01702bd7142f2c651cd1f7ca8dbe9aac2b602f5e56778","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-09T16:02:23Z","title_canon_sha256":"05ac7fb5cd4da151cfb073a70370c9c2d698939591d5493cd8f040819911cc64"},"schema_version":"1.0","source":{"id":"2505.06150","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.06150","created_at":"2026-07-05T11:14:28Z"},{"alias_kind":"arxiv_version","alias_value":"2505.06150v2","created_at":"2026-07-05T11:14:28Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.06150","created_at":"2026-07-05T11:14:28Z"},{"alias_kind":"pith_short_12","alias_value":"QTJGAUGHWPJF","created_at":"2026-07-05T11:14:28Z"},{"alias_kind":"pith_short_16","alias_value":"QTJGAUGHWPJF7TS6","created_at":"2026-07-05T11:14:28Z"},{"alias_kind":"pith_short_8","alias_value":"QTJGAUGH","created_at":"2026-07-05T11:14:28Z"}],"graph_snapshots":[{"event_id":"sha256:dce11f14ddbe18fff1a54e04ee204d1d09fd4191b6d805d2e139e22ffc4cfcfa","target":"graph","created_at":"2026-07-05T11:14:28Z","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/2505.06150/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We introduce a scaling law for fine-tuning large language models (LLMs) under fixed compute budgets that explicitly accounts for data composition. Conventional approaches measure training data solely by total tokens, yet the number of examples and their average token length -- what we term \\emph{dataset volume} -- play a decisive role in model performance. Our formulation is tuned following established procedures. Experiments on the BRICC dataset \\cite{salavati2024reducing} and subsets of the MMLU dataset \\cite{hendrycks2021measuringmassivemultitasklanguage}, evaluated under multiple subsampli","authors_text":"Aidan Kierans, Avijit Ghosh, Ryan Lagasse, Shiri Dori-Hacohen","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-09T16:02:23Z","title":"A Scaling Law for Token Efficiency in LLM Fine-Tuning Under Fixed Compute Budgets"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.06150","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":"sha256:5fa46d06f7d0cb38c5d688bbb84ce5fe26b435d0627029a008faf00639415e13","target":"record","created_at":"2026-07-05T11:14:28Z","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":"9bd95205037c05b650e01702bd7142f2c651cd1f7ca8dbe9aac2b602f5e56778","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-09T16:02:23Z","title_canon_sha256":"05ac7fb5cd4da151cfb073a70370c9c2d698939591d5493cd8f040819911cc64"},"schema_version":"1.0","source":{"id":"2505.06150","kind":"arxiv","version":2}},"canonical_sha256":"84d26050c7b3d25fce5e03b884041b8bd814e1fe582a64501429c030261335ca","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"84d26050c7b3d25fce5e03b884041b8bd814e1fe582a64501429c030261335ca","first_computed_at":"2026-07-05T11:14:28.919373Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:14:28.919373Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Z26b7CGb7eiU3/DNIK2Dx2nfwESEgs7f9uFcdIFdBUEuesQkS06owWj6GsU4vGszjENgi/uYm4fm50gLUsOjDw==","signature_status":"signed_v1","signed_at":"2026-07-05T11:14:28.919851Z","signed_message":"canonical_sha256_bytes"},"source_id":"2505.06150","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:5fa46d06f7d0cb38c5d688bbb84ce5fe26b435d0627029a008faf00639415e13","sha256:dce11f14ddbe18fff1a54e04ee204d1d09fd4191b6d805d2e139e22ffc4cfcfa"],"state_sha256":"fea141fa29a211c1da509ae55ca507418c46b687c1bc4fe8311876857629dd36"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"i2fQHRSmDo6neuAx6VPfhlCcMrn5inFR4ZZ5HLBdEl1sYJx2vEHQrEFKlFZrOsx6XI6Xc8/bN2Ikk5R9BNXXAw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-23T22:00:31.472113Z","bundle_sha256":"7aacb6e3ab566a36959ec0e37ce388cce5355309d3ca56494e87dbd1ca50094f"}}