{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:DOSG6QPNY3PB3ZCKVYM75G5TDA","short_pith_number":"pith:DOSG6QPN","canonical_record":{"source":{"id":"2505.21930","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-05-28T03:27:08Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"7a7223b9262aa90b3c703faa5aac082e08951f431b78134be2b7d5192ec759b8","abstract_canon_sha256":"5394df00ad4e50dad23e7cf7a090733de277b92f909ed8131e0f7bd5b24958a2"},"schema_version":"1.0"},"canonical_sha256":"1ba46f41edc6de1de44aae19fe9bb318325e19e5f20af1937651a57a3c6a20f2","source":{"kind":"arxiv","id":"2505.21930","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.21930","created_at":"2026-07-05T11:11:06Z"},{"alias_kind":"arxiv_version","alias_value":"2505.21930v1","created_at":"2026-07-05T11:11:06Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.21930","created_at":"2026-07-05T11:11:06Z"},{"alias_kind":"pith_short_12","alias_value":"DOSG6QPNY3PB","created_at":"2026-07-05T11:11:06Z"},{"alias_kind":"pith_short_16","alias_value":"DOSG6QPNY3PB3ZCK","created_at":"2026-07-05T11:11:06Z"},{"alias_kind":"pith_short_8","alias_value":"DOSG6QPN","created_at":"2026-07-05T11:11:06Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:DOSG6QPNY3PB3ZCKVYM75G5TDA","target":"record","payload":{"canonical_record":{"source":{"id":"2505.21930","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-05-28T03:27:08Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"7a7223b9262aa90b3c703faa5aac082e08951f431b78134be2b7d5192ec759b8","abstract_canon_sha256":"5394df00ad4e50dad23e7cf7a090733de277b92f909ed8131e0f7bd5b24958a2"},"schema_version":"1.0"},"canonical_sha256":"1ba46f41edc6de1de44aae19fe9bb318325e19e5f20af1937651a57a3c6a20f2","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:11:06.123296Z","signature_b64":"hoXQIjEa+oyC3xt3U+3DjdTDS3MBc085x1sX8k9FzzBcUXAGzWBzaIDXGilIoE6isUm78vFnJcpDsC0JMHwEAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1ba46f41edc6de1de44aae19fe9bb318325e19e5f20af1937651a57a3c6a20f2","last_reissued_at":"2026-07-05T11:11:06.122884Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:11:06.122884Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2505.21930","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:11:06Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"8dCVDngXk4Qp6A1MsokjOyJ5GC6YbctUax9IsJDbUDucqEfb/LFrKpval6St6iT5ia6oSgDXenDRrBtHktyRCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T13:15:33.376211Z"},"content_sha256":"45f00bad9e386525277310cddcdb8ac25b30510bb73525f17520e939b28c72d3","schema_version":"1.0","event_id":"sha256:45f00bad9e386525277310cddcdb8ac25b30510bb73525f17520e939b28c72d3"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:DOSG6QPNY3PB3ZCKVYM75G5TDA","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Efficient Ensemble for Fine-tuning Language Models on Multiple Datasets","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Dongyue Li, Hongyang R. Zhang, Lu Wang, Ziniu Zhang","submitted_at":"2025-05-28T03:27:08Z","abstract_excerpt":"This paper develops an ensemble method for fine-tuning a language model to multiple datasets. Existing methods, such as quantized LoRA (QLoRA), are efficient when adapting to a single dataset. When training on multiple datasets of different tasks, a common setup in practice, it remains unclear how to design an efficient adaptation for fine-tuning language models. We propose to use an ensemble of multiple smaller adapters instead of a single adapter per task. We design an efficient algorithm that partitions $n$ datasets into $m$ groups, where $m$ is typically much smaller than $n$ in practice, "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.21930","kind":"arxiv","version":1},"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/2505.21930/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":"2026-07-05T11:11:06Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"iQGTESX/knSHehrg1nLhLyeu9hdW3+tqztjOUjLtIYCEOS8n7nIE3AoXTdjo/QsYfyyTDM2ve1OmGciS3XIGBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T13:15:33.376723Z"},"content_sha256":"72f4240ef96104d29d9a5e33edeeda3cff9a001074d68ef08e61425cf17b20c2","schema_version":"1.0","event_id":"sha256:72f4240ef96104d29d9a5e33edeeda3cff9a001074d68ef08e61425cf17b20c2"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/DOSG6QPNY3PB3ZCKVYM75G5TDA/bundle.json","state_url":"https://pith.science/pith/DOSG6QPNY3PB3ZCKVYM75G5TDA/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/DOSG6QPNY3PB3ZCKVYM75G5TDA/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":"2026-08-08T13:15:33Z","links":{"resolver":"https://pith.science/pith/DOSG6QPNY3PB3ZCKVYM75G5TDA","bundle":"https://pith.science/pith/DOSG6QPNY3PB3ZCKVYM75G5TDA/bundle.json","state":"https://pith.science/pith/DOSG6QPNY3PB3ZCKVYM75G5TDA/state.json","well_known_bundle":"https://pith.science/.well-known/pith/DOSG6QPNY3PB3ZCKVYM75G5TDA/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:DOSG6QPNY3PB3ZCKVYM75G5TDA","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":"5394df00ad4e50dad23e7cf7a090733de277b92f909ed8131e0f7bd5b24958a2","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-05-28T03:27:08Z","title_canon_sha256":"7a7223b9262aa90b3c703faa5aac082e08951f431b78134be2b7d5192ec759b8"},"schema_version":"1.0","source":{"id":"2505.21930","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.21930","created_at":"2026-07-05T11:11:06Z"},{"alias_kind":"arxiv_version","alias_value":"2505.21930v1","created_at":"2026-07-05T11:11:06Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.21930","created_at":"2026-07-05T11:11:06Z"},{"alias_kind":"pith_short_12","alias_value":"DOSG6QPNY3PB","created_at":"2026-07-05T11:11:06Z"},{"alias_kind":"pith_short_16","alias_value":"DOSG6QPNY3PB3ZCK","created_at":"2026-07-05T11:11:06Z"},{"alias_kind":"pith_short_8","alias_value":"DOSG6QPN","created_at":"2026-07-05T11:11:06Z"}],"graph_snapshots":[{"event_id":"sha256:72f4240ef96104d29d9a5e33edeeda3cff9a001074d68ef08e61425cf17b20c2","target":"graph","created_at":"2026-07-05T11:11:06Z","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/2505.21930/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"This paper develops an ensemble method for fine-tuning a language model to multiple datasets. Existing methods, such as quantized LoRA (QLoRA), are efficient when adapting to a single dataset. When training on multiple datasets of different tasks, a common setup in practice, it remains unclear how to design an efficient adaptation for fine-tuning language models. We propose to use an ensemble of multiple smaller adapters instead of a single adapter per task. We design an efficient algorithm that partitions $n$ datasets into $m$ groups, where $m$ is typically much smaller than $n$ in practice, ","authors_text":"Dongyue Li, Hongyang R. Zhang, Lu Wang, Ziniu Zhang","cross_cats":["cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-05-28T03:27:08Z","title":"Efficient Ensemble for Fine-tuning Language Models on Multiple Datasets"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.21930","kind":"arxiv","version":1},"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":"sha256:45f00bad9e386525277310cddcdb8ac25b30510bb73525f17520e939b28c72d3","target":"record","created_at":"2026-07-05T11:11:06Z","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":"5394df00ad4e50dad23e7cf7a090733de277b92f909ed8131e0f7bd5b24958a2","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-05-28T03:27:08Z","title_canon_sha256":"7a7223b9262aa90b3c703faa5aac082e08951f431b78134be2b7d5192ec759b8"},"schema_version":"1.0","source":{"id":"2505.21930","kind":"arxiv","version":1}},"canonical_sha256":"1ba46f41edc6de1de44aae19fe9bb318325e19e5f20af1937651a57a3c6a20f2","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"1ba46f41edc6de1de44aae19fe9bb318325e19e5f20af1937651a57a3c6a20f2","first_computed_at":"2026-07-05T11:11:06.122884Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:11:06.122884Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"hoXQIjEa+oyC3xt3U+3DjdTDS3MBc085x1sX8k9FzzBcUXAGzWBzaIDXGilIoE6isUm78vFnJcpDsC0JMHwEAQ==","signature_status":"signed_v1","signed_at":"2026-07-05T11:11:06.123296Z","signed_message":"canonical_sha256_bytes"},"source_id":"2505.21930","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:45f00bad9e386525277310cddcdb8ac25b30510bb73525f17520e939b28c72d3","sha256:72f4240ef96104d29d9a5e33edeeda3cff9a001074d68ef08e61425cf17b20c2"],"state_sha256":"6d112d6c922d0a3d4eaf1c2b3c7f734ec7da373490ecf9a07bbf8ea0281e95f6"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"8nlYRgIJ/MUApMLqa48146Kgcn7HmaMgbDZAFJFTqRDHuE2/OQ9RQJRvDisrDhIt/9oTVsZocwWIXDa7rr+rDw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-08T13:15:33.381859Z","bundle_sha256":"f730f10cad8c7954cb7397ca2abb4639dc67c8d8a629da896d670cfdf21fe9a3"}}