{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/localize-and-stitch-efficient-model-merging","title":"Localize-and-Stitch: Efficient Model Merging via Sparse Task Arithmetic","arxiv_id":"2408.13656","date":"2024-08-24","proceeding":null,"authors":["Yifei He","Yuzheng Hu","Yong Lin","Tong Zhang","Han Zhao"],"abstract":"Model merging offers an effective strategy to combine the strengths of multiple finetuned models into a unified model that preserves the specialized capabilities of each. Existing methods merge models in a global manner, performing arithmetic operations across all model parameters. However, such global merging often leads to task interference, degrading the performance of the merged model. In this work, we introduce Localize-and-Stitch, a novel approach that merges models in a localized way. Our algorithm works in two steps: i) Localization: identify tiny ($1\\%$ of the total parameters) localized regions in the finetuned models containing essential skills for the downstream tasks, and ii) Stitching: reintegrate only these essential regions back into the pretrained model for task synergy. We demonstrate that our approach effectively locates sparse regions responsible for finetuned performance, and the localized regions could be treated as compact and interpretable representations of the finetuned models (tasks). Empirically, we evaluate our method on various vision and language benchmarks, showing that it outperforms existing model merging methods under different data availability scenarios. Beyond strong empirical performance, our algorithm also facilitates model compression and preserves pretrained knowledge, enabling flexible and continual skill composition from multiple finetuned models with minimal storage and computational overhead. Our code is available at https://github.com/uiuctml/Localize-and-Stitch.","url_abs":"https://arxiv.org/abs/2408.13656v2","url_pdf":"https://arxiv.org/pdf/2408.13656v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"localize-and-stitch-efficient-model-merging","repo_url":"https://github.com/uiuctml/localize-and-stitch","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"localize-and-stitch-efficient-model-merging","repo_url":"https://github.com/yifei-he/localize-and-stitch","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"model-compression","task_name":"Model Compression"},{"task_slug":"task-arithmetic","task_name":"Task Arithmetic"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2408.13656","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2408.13656"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/uiuctml/localize-and-stitch","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/yifei-he/localize-and-stitch","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran_draft_wrong":2,"ran":5},"by_repo_kind":{"official":{"samples":7,"ran":7,"repositories":2}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"3053d2bdc3ea5c89","entry":"create_log_dir","repo":"yifei-he/localize-and-stitch","repo_kind":"official","path":"vision/src/main_dataless_localize_stitch.py","file_url":"https://github.com/yifei-he/localize-and-stitch/blob/HEAD/vision/src/main_dataless_localize_stitch.py","link_basis":"plan_row","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"3053d2bdc3ea5c89"}},{"code_sha256_prefix":"e768e25ca95f70cd","entry":"input_example_to_string","repo":"yifei-he/localize-and-stitch","repo_kind":"official","path":"language/src/dataset.py","file_url":"https://github.com/yifei-he/localize-and-stitch/blob/HEAD/language/src/dataset.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"e768e25ca95f70cd"}},{"code_sha256_prefix":"fbef70dc7c700501","entry":"input_example_to_tuple","repo":"yifei-he/localize-and-stitch","repo_kind":"official","path":"language/src/dataset.py","file_url":"https://github.com/yifei-he/localize-and-stitch/blob/HEAD/language/src/dataset.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"fbef70dc7c700501"}},{"code_sha256_prefix":"b2160a3a2b890e58","entry":"parse_data_arguments","repo":"yifei-he/localize-and-stitch","repo_kind":"official","path":"language/args.py","file_url":"https://github.com/yifei-he/localize-and-stitch/blob/HEAD/language/args.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"b2160a3a2b890e58"}},{"code_sha256_prefix":"883ebc0d4d55db02","entry":"select_trainable_parameters","repo":"uiuctml/localize-and-stitch","repo_kind":"official","path":"language/main_localize_stitch.py","file_url":"https://github.com/uiuctml/localize-and-stitch/blob/HEAD/language/main_localize_stitch.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"883ebc0d4d55db02"}},{"code_sha256_prefix":"b62076c5483738dc","entry":"text_classification_metrics","repo":"yifei-he/localize-and-stitch","repo_kind":"official","path":"language/src/processors.py","file_url":"https://github.com/yifei-he/localize-and-stitch/blob/HEAD/language/src/processors.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"b62076c5483738dc"}},{"code_sha256_prefix":"e3eed9fd49cd08fb","entry":"tokenize_multipart_input","repo":"yifei-he/localize-and-stitch","repo_kind":"official","path":"language/src/dataset.py","file_url":"https://github.com/yifei-he/localize-and-stitch/blob/HEAD/language/src/dataset.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"e3eed9fd49cd08fb"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}