{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:ENTSYT6WTBYWL65J2P4GOVRD6W","short_pith_number":"pith:ENTSYT6W","schema_version":"1.0","canonical_sha256":"23672c4fd6987165fba9d3f8675623f582822b6b7de83437d51cc402807008ba","source":{"kind":"arxiv","id":"2410.13841","version":1},"attestation_state":"computed","paper":{"title":"A Unified View of Delta Parameter Editing in Post-Trained Large-Scale Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Bowen Yu, Hongyu Lin, Keming Lu, Le Sun, Le Yu, Qiaoyu Tang, Xianpei Han, Yaojie Lu","submitted_at":"2024-10-17T17:56:53Z","abstract_excerpt":"Post-training has emerged as a crucial paradigm for adapting large-scale pre-trained models to various tasks, whose effects are fully reflected by delta parameters (i.e., the disparity between post-trained and pre-trained parameters). While numerous studies have explored delta parameter properties via operations like pruning, quantization, low-rank approximation, and extrapolation, a unified framework for systematically examining these characteristics has been lacking. In this paper, we propose a novel perspective based on Riemann sum approximation of the loss function to elucidate delta param"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2410.13841","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-10-17T17:56:53Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"dba3c23666ef5836153ad9a5956149419334fef74578db9210d560e4823ea76c","abstract_canon_sha256":"691e9152c06d50d1e823cf29fedd0a8d36bef42d6a71c2602eac91450bf7abb2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:22:06.405749Z","signature_b64":"AX5iVxoXeGp4LGYGRplZ2ARL1tSnJLYWf67j00i3cDnfrqVGZ8Ao4giTrQ7SZcaA0rSKKf1JjR0s1Q7tkF+OCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"23672c4fd6987165fba9d3f8675623f582822b6b7de83437d51cc402807008ba","last_reissued_at":"2026-07-05T09:22:06.405264Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:22:06.405264Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"A Unified View of Delta Parameter Editing in Post-Trained Large-Scale Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Bowen Yu, Hongyu Lin, Keming Lu, Le Sun, Le Yu, Qiaoyu Tang, Xianpei Han, Yaojie Lu","submitted_at":"2024-10-17T17:56:53Z","abstract_excerpt":"Post-training has emerged as a crucial paradigm for adapting large-scale pre-trained models to various tasks, whose effects are fully reflected by delta parameters (i.e., the disparity between post-trained and pre-trained parameters). While numerous studies have explored delta parameter properties via operations like pruning, quantization, low-rank approximation, and extrapolation, a unified framework for systematically examining these characteristics has been lacking. In this paper, we propose a novel perspective based on Riemann sum approximation of the loss function to elucidate delta param"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.13841","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/2410.13841/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2410.13841","created_at":"2026-07-05T09:22:06.405324+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.13841v1","created_at":"2026-07-05T09:22:06.405324+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.13841","created_at":"2026-07-05T09:22:06.405324+00:00"},{"alias_kind":"pith_short_12","alias_value":"ENTSYT6WTBYW","created_at":"2026-07-05T09:22:06.405324+00:00"},{"alias_kind":"pith_short_16","alias_value":"ENTSYT6WTBYWL65J","created_at":"2026-07-05T09:22:06.405324+00:00"},{"alias_kind":"pith_short_8","alias_value":"ENTSYT6W","created_at":"2026-07-05T09:22:06.405324+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.01048","citing_title":"Compared to What? Baselines and Metrics for Counterfactual Prompting","ref_index":75,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ENTSYT6WTBYWL65J2P4GOVRD6W","json":"https://pith.science/pith/ENTSYT6WTBYWL65J2P4GOVRD6W.json","graph_json":"https://pith.science/api/pith-number/ENTSYT6WTBYWL65J2P4GOVRD6W/graph.json","events_json":"https://pith.science/api/pith-number/ENTSYT6WTBYWL65J2P4GOVRD6W/events.json","paper":"https://pith.science/paper/ENTSYT6W"},"agent_actions":{"view_html":"https://pith.science/pith/ENTSYT6WTBYWL65J2P4GOVRD6W","download_json":"https://pith.science/pith/ENTSYT6WTBYWL65J2P4GOVRD6W.json","view_paper":"https://pith.science/paper/ENTSYT6W","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.13841&json=true","fetch_graph":"https://pith.science/api/pith-number/ENTSYT6WTBYWL65J2P4GOVRD6W/graph.json","fetch_events":"https://pith.science/api/pith-number/ENTSYT6WTBYWL65J2P4GOVRD6W/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ENTSYT6WTBYWL65J2P4GOVRD6W/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ENTSYT6WTBYWL65J2P4GOVRD6W/action/storage_attestation","attest_author":"https://pith.science/pith/ENTSYT6WTBYWL65J2P4GOVRD6W/action/author_attestation","sign_citation":"https://pith.science/pith/ENTSYT6WTBYWL65J2P4GOVRD6W/action/citation_signature","submit_replication":"https://pith.science/pith/ENTSYT6WTBYWL65J2P4GOVRD6W/action/replication_record"}},"created_at":"2026-07-05T09:22:06.405324+00:00","updated_at":"2026-07-05T09:22:06.405324+00:00"}