{"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/disperse-then-merge-pushing-the-limits-of","title":"Disperse-Then-Merge: Pushing the Limits of Instruction Tuning via Alignment Tax Reduction","arxiv_id":"2405.13432","date":"2024-05-22","proceeding":null,"authors":["Tingchen Fu","Deng Cai","Lemao Liu","Shuming Shi","Rui Yan"],"abstract":"Supervised fine-tuning (SFT) on instruction-following corpus is a crucial approach toward the alignment of large language models (LLMs). However, the performance of LLMs on standard knowledge and reasoning benchmarks tends to suffer from deterioration at the latter stage of the SFT process, echoing the phenomenon of alignment tax. Through our pilot study, we put a hypothesis that the data biases are probably one cause behind the phenomenon. To address the issue, we introduce a simple disperse-then-merge framework. To be concrete, we disperse the instruction-following data into portions and train multiple sub-models using different data portions. Then we merge multiple models into a single one via model merging techniques. Despite its simplicity, our framework outperforms various sophisticated methods such as data curation and training regularization on a series of standard knowledge and reasoning benchmarks.","url_abs":"https://arxiv.org/abs/2405.13432v1","url_pdf":"https://arxiv.org/pdf/2405.13432v1.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":"disperse-then-merge-pushing-the-limits-of","repo_url":"https://github.com/TingchenFu/ACL24-ExpertFusion","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"instruction-following","task_name":"Instruction Following"}],"methods":[{"method_slug":"sft","method_name":"SFT"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2405.13432","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2405.13432"}},"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/TingchenFu/ACL24-ExpertFusion","reach":{"status":"ok"}}],"summary":{"ran":2,"unverified":1},"by_repo_kind":{"official":{"samples":3,"ran":2,"repositories":1}},"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":3,"samples":[{"code_sha256_prefix":"beaf421b457284c5","entry":"aggregate","repo":"TingchenFu/ACL24-ExpertFusion","repo_kind":"official","path":"code/merge_utils.py","file_url":"https://github.com/TingchenFu/ACL24-ExpertFusion/blob/HEAD/code/merge_utils.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"beaf421b457284c5"}},{"code_sha256_prefix":"06a81d10e4fb3a93","entry":"vector2statedict","repo":"TingchenFu/ACL24-ExpertFusion","repo_kind":"official","path":"code/merge_utils.py","file_url":"https://github.com/TingchenFu/ACL24-ExpertFusion/blob/HEAD/code/merge_utils.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"06a81d10e4fb3a93"}},{"code_sha256_prefix":"c2301026042bef4c","entry":"statedict2vector","repo":"TingchenFu/ACL24-ExpertFusion","repo_kind":"official","path":"code/merge_utils.py","file_url":"https://github.com/TingchenFu/ACL24-ExpertFusion/blob/HEAD/code/merge_utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"c2301026042bef4c"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}