{"about":{"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.","site":"https://codewithpapers.app","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","syntology":{"site":"https://syntology.ai","developers":"https://syntology.ai/developers","mcp":{"server":"https://syntology.ai/mcp","transport":"streamable-http","server_card":"https://syntology.ai/.well-known/mcp/server-card.json","auth":{"type":"trial token, no account","trial_token":"https://syntology.ai/api/oauth/trial/token","method":"POST","docs":"https://syntology.ai/developers"}},"have":"https://syntology.ai/api/graph/have?x=<method, arXiv id or title> (free, answers coverage only)","paper_base":"https://syntology.ai/paper/","atlas_base":"https://app.syntology.ai/?focus="},"machine_readable":[{"url":"https://codewithpapers.app/llms.txt","what":"the machine catalog: every machine-readable file, counted"},{"url":"https://codewithpapers.app/index/manifest.json","what":"paper-to-code index by arXiv id, with Syntology's counts"},{"url":"https://codewithpapers.app/search/manifest.json","what":"site search index (titles, authors) and its files"},{"url":"https://codewithpapers.app/download","what":"bulk files: Syntology's layer, described there"},{"url":"https://codewithpapers.app/build_manifest.json","what":"the build record: inputs, counts, exclusions, probes"}]},"url":"/task/task-arithmetic","name":"Task Arithmetic","slug":"task-arithmetic","description_markdown":"A task vector specifies a direction in the weight space of a pre-trained\r\nmodel, such that movement in that direction improves performance on the task. We build task vectors by subtracting the weights of a pre-trained model from the weights of the same model after fine-tuning on a task. We show that these task vectors can be modified and combined together through arithmetic operations such as negation and addition, and the behavior of the resulting model is steered accordingly.","categories":[{"name":"Methodology","url":"/area/methodology"}],"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","slug_source":"derived"},"counts":{"papers_tagged":61,"papers_with_code":24,"benchmarks":0,"benchmark_tables_in_archive":0,"benchmark_tables_shown":0,"benchmark_tables_withheld_as_spam":0,"benchmark_definition":"a leaderboard table with at least one row; benchmark_tables_shown also counts the zero-row tables; benchmark_tables_in_archive adds the tables withheld as spam","datasets":0,"subtasks":0,"parent_tasks":1},"benchmarks":[],"datasets":[],"subtasks":[],"parent_tasks":[{"url":"/task/multi-task-learning","name":"Multi-Task Learning"}],"papers":{"order":"repositories listed in the archive (most first), then date (newest first), then slug; the archive holds no stars","population":"every paper the archive tags with this task (distinct papers; spam titles not listed). 'of' counts the papers listed; until 2026-09-25 it counted only the papers with a code link, and that number is of_with_code","page":1,"pages_in_order":1,"rows_per_page":100,"shown":61,"of":61,"of_with_code":24,"tagged_in_all":61,"counts":{"archive_papers_tagged":61,"with_a_code_link":24,"where_syntology_ran_a_sample":18,"not_listed_spam_title":0,"listed":61,"listed_where_code_ran":18,"where_syntology_ran_a_sample_split":{"with_a_run_with_no_instrument_failure":14,"every_run_a_failure_of_syntologys_instrument":4,"listed_with_a_run_with_no_instrument_failure":14,"listed_every_run_a_failure_of_syntologys_instrument":4,"filter":{"states":["a run with no instrument failure","any run, instrument failures included"],"default":"a run with no instrument failure","note":"on the 'only where code ran' pages the default hides, in the browser, the rows where every run was a failure of Syntology's instrument; the second state shows them again. Rows are hidden, never re-ordered; these twins list every row"}},"definition":"distinct papers the archive tags; 'where Syntology ran a sample' counts papers with at least one harvested sample that ran, which is not a correctness claim"},"items":[{"url":"/paper/editing-models-with-task-arithmetic","slug":"editing-models-with-task-arithmetic","title":"Editing Models with Task Arithmetic","date":"2022-12-08","arxiv_id":"2212.04089","repositories_listed":9,"syntology":{"n":15,"n_ran":5,"n_constructed":3,"n_ran_checked":3,"n_instrument":2,"n_unverified":10,"n_honours":0,"n_violates":0,"n_no_contract":3,"n_pointer_only":4,"phrase":"5 ran (of which 3 constructed an object rather than computing a result; 3 with no instrument failure: 0 honoured, 0 violated, 3 with no contract checked; 2 where Syntology's instrument failed) · 10 unverified","sample_list":"/paper/editing-models-with-task-arithmetic#ran","syntology_url":"https://syntology.ai/paper/2212.04089","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2212.04089"}},"official":{"repos":["mlfoundations/task_vectors"],"state":"official: no sample here; runs from other or unrecorded repositories","n_ran":0,"n_constructed":0,"n_ran_no_instrument_failure":0,"n_unverified":0,"ran_from_kinds":["listed","unlocated"]}}},{"url":"/paper/language-models-are-homer-simpson-safety-re","slug":"language-models-are-homer-simpson-safety-re","title":"Language Models are Homer Simpson! Safety Re-Alignment of Fine-tuned Language Models through Task Arithmetic","date":"2024-02-19","arxiv_id":"2402.11746","repositories_listed":3,"syntology":{"n":16,"n_ran":13,"n_constructed":0,"n_ran_checked":2,"n_instrument":11,"n_unverified":3,"n_honours":1,"n_violates":1,"n_no_contract":0,"n_pointer_only":9,"phrase":"13 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 1 honoured, 1 violated, 0 with no contract checked; 11 where Syntology's instrument failed) · 3 unverified","sample_list":"/paper/language-models-are-homer-simpson-safety-re#ran","syntology_url":"https://syntology.ai/paper/2402.11746","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2402.11746"}},"official":{"repos":["declare-lab/resta","hiyouga/llama-factory"],"state":"official (archive's flag): 6 ran","n_ran":6,"n_constructed":0,"n_ran_no_instrument_failure":2,"n_unverified":1,"ran_from_kinds":["listed","official"]}}},{"url":"/paper/localize-and-stitch-efficient-model-merging","slug":"localize-and-stitch-efficient-model-merging","title":"Localize-and-Stitch: Efficient Model Merging via Sparse Task Arithmetic","date":"2024-08-24","arxiv_id":"2408.13656","repositories_listed":2,"syntology":{"n":7,"n_ran":7,"n_constructed":0,"n_ran_checked":2,"n_instrument":5,"n_unverified":0,"n_honours":0,"n_violates":0,"n_no_contract":2,"n_pointer_only":0,"phrase":"7 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 5 where Syntology's instrument failed) · 0 unverified","sample_list":"/paper/localize-and-stitch-efficient-model-merging#ran","syntology_url":"https://syntology.ai/paper/2408.13656","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2408.13656"}},"official":{"repos":["uiuctml/localize-and-stitch","yifei-he/localize-and-stitch"],"state":"official (archive's flag): 7 ran","n_ran":7,"n_constructed":0,"n_ran_no_instrument_failure":2,"n_unverified":0,"ran_from_kinds":["official"]}}},{"url":"/paper/fine-tuning-linear-layers-only-is-a-simple","slug":"fine-tuning-linear-layers-only-is-a-simple","title":"Fine-Tuning Attention Modules Only: Enhancing Weight Disentanglement in Task Arithmetic","date":"2024-07-09","arxiv_id":"2407.07089","repositories_listed":2,"syntology":{"n":5,"n_ran":3,"n_constructed":0,"n_ran_checked":0,"n_instrument":3,"n_unverified":2,"n_honours":0,"n_violates":0,"n_no_contract":0,"n_pointer_only":0,"phrase":"3 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 3 where Syntology's instrument failed) · 2 unverified","sample_list":"/paper/fine-tuning-linear-layers-only-is-a-simple#ran","syntology_url":"https://syntology.ai/paper/2407.07089","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2407.07089"}},"official":{"repos":["kyrie-23/task_arithmetic_tangent","kyrie-23/linear_task_arithmetic"],"state":"official (archive's flag): 3 ran","n_ran":3,"n_constructed":0,"n_ran_no_instrument_failure":0,"n_unverified":2,"ran_from_kinds":["official"]}}},{"url":"/paper/localizing-task-information-for-improved","slug":"localizing-task-information-for-improved","title":"Localizing Task Information for Improved Model Merging and Compression","date":"2024-05-13","arxiv_id":"2405.07813","repositories_listed":2,"syntology":{"n":5,"n_ran":3,"n_constructed":0,"n_ran_checked":0,"n_instrument":3,"n_unverified":2,"n_honours":0,"n_violates":0,"n_no_contract":0,"n_pointer_only":4,"phrase":"3 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 3 where Syntology's instrument failed) · 2 unverified","sample_list":"/paper/localizing-task-information-for-improved#ran","syntology_url":"https://syntology.ai/paper/2405.07813","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2405.07813"}},"official":{"repos":["nik-dim/tall_masks"],"state":"official (archive's flag): 3 ran","n_ran":3,"n_constructed":0,"n_ran_no_instrument_failure":0,"n_unverified":2,"ran_from_kinds":["official"]}}},{"url":"/paper/calm-consensus-aware-localized-merging-for","slug":"calm-consensus-aware-localized-merging-for","title":"CALM: Consensus-Aware Localized Merging for Multi-Task Learning","date":"2025-06-16","arxiv_id":"2506.13406","repositories_listed":1,"syntology":{"n":10,"n_ran":6,"n_constructed":0,"n_ran_checked":3,"n_instrument":3,"n_unverified":4,"n_honours":1,"n_violates":0,"n_no_contract":2,"n_pointer_only":10,"phrase":"6 ran (of which 0 constructed an object rather than computing a result; 3 with no instrument failure: 1 honoured, 0 violated, 2 with no contract checked; 3 where Syntology's instrument failed) · 4 unverified","sample_list":"/paper/calm-consensus-aware-localized-merging-for#ran","syntology_url":"https://syntology.ai/paper/2506.13406","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2506.13406"}},"official":{"repos":["yankd22/calm"],"state":"official (archive's flag): 6 ran","n_ran":6,"n_constructed":0,"n_ran_no_instrument_failure":3,"n_unverified":4,"ran_from_kinds":["official"]}}},{"url":"/paper/cross-model-transfer-of-task-vectors-via-few","slug":"cross-model-transfer-of-task-vectors-via-few","title":"Cross-Model Transfer of Task Vectors via Few-Shot Orthogonal Alignment","date":"2025-05-17","arxiv_id":"2505.12021","repositories_listed":1,"syntology":null},{"url":"/paper/investigating-task-arithmetic-for-zero-shot","slug":"investigating-task-arithmetic-for-zero-shot","title":"Investigating Task Arithmetic for Zero-Shot Information Retrieval","date":"2025-05-01","arxiv_id":"2505.00649","repositories_listed":1,"syntology":null},{"url":"/paper/leveraging-submodule-linearity-enhances-task","slug":"leveraging-submodule-linearity-enhances-task","title":"Leveraging Submodule Linearity Enhances Task Arithmetic Performance in LLMs","date":"2025-04-15","arxiv_id":"2504.10902","repositories_listed":1,"syntology":{"n":1,"n_ran":1,"n_constructed":0,"n_ran_checked":1,"n_instrument":0,"n_unverified":0,"n_honours":1,"n_violates":0,"n_no_contract":0,"n_pointer_only":1,"phrase":"1 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 1 honoured, 0 violated, 0 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified","sample_list":"/paper/leveraging-submodule-linearity-enhances-task#ran","syntology_url":"https://syntology.ai/paper/2504.10902","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2504.10902"}},"official":{"repos":["deep-analysis-research/slta"],"state":"official (archive's flag): 1 ran","n_ran":1,"n_constructed":0,"n_ran_no_instrument_failure":1,"n_unverified":0,"ran_from_kinds":["official"]}}},{"url":"/paper/efficient-model-editing-with-task-localized","slug":"efficient-model-editing-with-task-localized","title":"Efficient Model Editing with Task-Localized Sparse Fine-tuning","date":"2025-04-03","arxiv_id":"2504.02620","repositories_listed":1,"syntology":null},{"url":"/paper/efficient-model-editing-with-task-vector","slug":"efficient-model-editing-with-task-vector","title":"Efficient Model Editing with Task Vector Bases: A Theoretical Framework and Scalable Approach","date":"2025-02-03","arxiv_id":"2502.01015","repositories_listed":1,"syntology":{"n":4,"n_ran":2,"n_constructed":0,"n_ran_checked":2,"n_instrument":0,"n_unverified":2,"n_honours":0,"n_violates":0,"n_no_contract":2,"n_pointer_only":2,"phrase":"2 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 0 where Syntology's instrument failed) · 2 unverified","sample_list":"/paper/efficient-model-editing-with-task-vector#ran","syntology_url":"https://syntology.ai/paper/2502.01015","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2502.01015"}},"official":{"repos":["uiuctml/TaskVectorBasis"],"state":"official (archive's flag): 2 ran","n_ran":2,"n_constructed":0,"n_ran_no_instrument_failure":2,"n_unverified":2,"ran_from_kinds":["official"]}}},{"url":"/paper/task-singular-vectors-reducing-task","slug":"task-singular-vectors-reducing-task","title":"Task Singular Vectors: Reducing Task Interference in Model Merging","date":"2024-11-26","arxiv_id":"2412.00081","repositories_listed":1,"syntology":null},{"url":"/paper/negmerge-consensual-weight-negation-for","slug":"negmerge-consensual-weight-negation-for","title":"NegMerge: Consensual Weight Negation for Strong Machine Unlearning","date":"2024-10-08","arxiv_id":"2410.05583","repositories_listed":1,"syntology":{"n":6,"n_ran":3,"n_constructed":0,"n_ran_checked":2,"n_instrument":1,"n_unverified":3,"n_honours":0,"n_violates":0,"n_no_contract":2,"n_pointer_only":6,"phrase":"3 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 1 where Syntology's instrument failed) · 3 unverified","sample_list":"/paper/negmerge-consensual-weight-negation-for#ran","syntology_url":"https://syntology.ai/paper/2410.05583","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2410.05583"}},"official":{"repos":["naver-ai/negmerge"],"state":"official (archive's flag): 3 ran","n_ran":3,"n_constructed":0,"n_ran_no_instrument_failure":2,"n_unverified":3,"ran_from_kinds":["official"]}}},{"url":"/paper/towards-diverse-device-heterogeneous","slug":"towards-diverse-device-heterogeneous","title":"Towards Diverse Device Heterogeneous Federated Learning via Task Arithmetic Knowledge Integration","date":"2024-09-27","arxiv_id":"2409.18461","repositories_listed":1,"syntology":null},{"url":"/paper/knowledge-composition-using-task-vectors-with","slug":"knowledge-composition-using-task-vectors-with","title":"Knowledge Composition using Task Vectors with Learned Anisotropic Scaling","date":"2024-07-03","arxiv_id":"2407.02880","repositories_listed":1,"syntology":{"n":4,"n_ran":3,"n_constructed":0,"n_ran_checked":1,"n_instrument":2,"n_unverified":1,"n_honours":0,"n_violates":1,"n_no_contract":0,"n_pointer_only":4,"phrase":"3 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 1 violated, 0 with no contract checked; 2 where Syntology's instrument failed) · 1 unverified","sample_list":"/paper/knowledge-composition-using-task-vectors-with#ran","syntology_url":"https://syntology.ai/paper/2407.02880","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2407.02880"}},"official":{"repos":["fredzzhang/atlas"],"state":"official (archive's flag): 3 ran","n_ran":3,"n_constructed":0,"n_ran_no_instrument_failure":1,"n_unverified":1,"ran_from_kinds":["official"]}}},{"url":"/paper/no-train-but-gain-language-arithmetic-for","slug":"no-train-but-gain-language-arithmetic-for","title":"No Train but Gain: Language Arithmetic for training-free Language Adapters enhancement","date":"2024-04-24","arxiv_id":"2404.15737","repositories_listed":1,"syntology":null},{"url":"/paper/have-you-merged-my-model-on-the-robustness-of","slug":"have-you-merged-my-model-on-the-robustness-of","title":"Have You Merged My Model? On The Robustness of Large Language Model IP Protection Methods Against Model Merging","date":"2024-04-08","arxiv_id":"2404.05188","repositories_listed":1,"syntology":{"n":1,"n_ran":1,"n_constructed":0,"n_ran_checked":1,"n_instrument":0,"n_unverified":0,"n_honours":0,"n_violates":1,"n_no_contract":0,"n_pointer_only":1,"phrase":"1 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 1 violated, 0 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified","sample_list":"/paper/have-you-merged-my-model-on-the-robustness-of#ran","syntology_url":"https://syntology.ai/paper/2404.05188","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2404.05188"}},"official":{"repos":["thuccslab/mergeguard"],"state":"official (archive's flag): 1 ran","n_ran":1,"n_constructed":0,"n_ran_no_instrument_failure":1,"n_unverified":0,"ran_from_kinds":["official"]}}},{"url":"/paper/merging-multi-task-models-via-weight","slug":"merging-multi-task-models-via-weight","title":"Merging Multi-Task Models via Weight-Ensembling Mixture of Experts","date":"2024-02-01","arxiv_id":"2402.00433","repositories_listed":1,"syntology":{"n":1,"n_ran":1,"n_constructed":0,"n_ran_checked":1,"n_instrument":0,"n_unverified":0,"n_honours":1,"n_violates":0,"n_no_contract":0,"n_pointer_only":1,"phrase":"1 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 1 honoured, 0 violated, 0 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified","sample_list":"/paper/merging-multi-task-models-via-weight#ran","syntology_url":"https://syntology.ai/paper/2402.00433","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2402.00433"}},"official":{"repos":["tanganke/weight-ensembling_moe"],"state":"official (archive's flag): 1 ran","n_ran":1,"n_constructed":0,"n_ran_no_instrument_failure":1,"n_unverified":0,"ran_from_kinds":["official"]}}},{"url":"/paper/concrete-subspace-learning-based-interference","slug":"concrete-subspace-learning-based-interference","title":"Concrete Subspace Learning based Interference Elimination for Multi-task Model Fusion","date":"2023-12-11","arxiv_id":"2312.06173","repositories_listed":1,"syntology":{"n":9,"n_ran":8,"n_constructed":0,"n_ran_checked":5,"n_instrument":3,"n_unverified":1,"n_honours":1,"n_violates":0,"n_no_contract":4,"n_pointer_only":9,"phrase":"8 ran (of which 0 constructed an object rather than computing a result; 5 with no instrument failure: 1 honoured, 0 violated, 4 with no contract checked; 3 where Syntology's instrument failed) · 1 unverified","sample_list":"/paper/concrete-subspace-learning-based-interference#ran","syntology_url":"https://syntology.ai/paper/2312.06173","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2312.06173"}},"official":{"repos":["tanganke/subspace_fusion"],"state":"official (archive's flag): 8 ran","n_ran":8,"n_constructed":0,"n_ran_no_instrument_failure":5,"n_unverified":1,"ran_from_kinds":["official"]}}},{"url":"/paper/model-merging-by-uncertainty-based-gradient","slug":"model-merging-by-uncertainty-based-gradient","title":"Model Merging by Uncertainty-Based Gradient Matching","date":"2023-10-19","arxiv_id":"2310.12808","repositories_listed":1,"syntology":{"n":2,"n_ran":2,"n_constructed":0,"n_ran_checked":2,"n_instrument":0,"n_unverified":0,"n_honours":0,"n_violates":0,"n_no_contract":2,"n_pointer_only":2,"phrase":"2 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified","sample_list":"/paper/model-merging-by-uncertainty-based-gradient#ran","syntology_url":"https://syntology.ai/paper/2310.12808","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2310.12808"}},"official":{"repos":["ukplab/iclr2024-model-merging"],"state":"official (archive's flag): 2 ran","n_ran":2,"n_constructed":0,"n_ran_no_instrument_failure":2,"n_unverified":0,"ran_from_kinds":["official"]}}},{"url":"/paper/parameter-efficient-multi-task-model-fusion","slug":"parameter-efficient-multi-task-model-fusion","title":"Parameter Efficient Multi-task Model Fusion with Partial Linearization","date":"2023-10-07","arxiv_id":"2310.04742","repositories_listed":1,"syntology":{"n":4,"n_ran":3,"n_constructed":0,"n_ran_checked":2,"n_instrument":1,"n_unverified":1,"n_honours":0,"n_violates":0,"n_no_contract":2,"n_pointer_only":4,"phrase":"3 ran (of which 0 constructed an object rather than computing a result; 2 with no instrument failure: 0 honoured, 0 violated, 2 with no contract checked; 1 where Syntology's instrument failed) · 1 unverified","sample_list":"/paper/parameter-efficient-multi-task-model-fusion#ran","syntology_url":"https://syntology.ai/paper/2310.04742","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2310.04742"}},"official":{"repos":["tanganke/peta"],"state":"official (archive's flag): 3 ran","n_ran":3,"n_constructed":0,"n_ran_no_instrument_failure":2,"n_unverified":1,"ran_from_kinds":["official"]}}},{"url":"/paper/adamerging-adaptive-model-merging-for-multi","slug":"adamerging-adaptive-model-merging-for-multi","title":"AdaMerging: Adaptive Model Merging for Multi-Task Learning","date":"2023-10-04","arxiv_id":"2310.02575","repositories_listed":1,"syntology":{"n":3,"n_ran":3,"n_constructed":0,"n_ran_checked":0,"n_instrument":3,"n_unverified":0,"n_honours":0,"n_violates":0,"n_no_contract":0,"n_pointer_only":0,"phrase":"3 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 3 where Syntology's instrument failed) · 0 unverified","sample_list":"/paper/adamerging-adaptive-model-merging-for-multi#ran","syntology_url":"https://syntology.ai/paper/2310.02575","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2310.02575"}},"official":{"repos":["EnnengYang/AdaMerging"],"state":"official (archive's flag): 3 ran","n_ran":3,"n_constructed":0,"n_ran_no_instrument_failure":0,"n_unverified":0,"ran_from_kinds":["official"]}}},{"url":"/paper/task-arithmetic-in-the-tangent-space-improved","slug":"task-arithmetic-in-the-tangent-space-improved","title":"Task Arithmetic in the Tangent Space: Improved Editing of Pre-Trained Models","date":"2023-05-22","arxiv_id":"2305.12827","repositories_listed":1,"syntology":{"n":4,"n_ran":1,"n_constructed":0,"n_ran_checked":1,"n_instrument":0,"n_unverified":3,"n_honours":0,"n_violates":0,"n_no_contract":1,"n_pointer_only":3,"phrase":"1 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 0 where Syntology's instrument failed) · 3 unverified","sample_list":"/paper/task-arithmetic-in-the-tangent-space-improved#ran","syntology_url":"https://syntology.ai/paper/2305.12827","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2305.12827"}},"official":{"repos":["gortizji/tangent_task_arithmetic"],"state":"official (archive's flag): 1 ran","n_ran":1,"n_constructed":0,"n_ran_no_instrument_failure":1,"n_unverified":3,"ran_from_kinds":["official"]}}},{"url":"/paper/an-empirical-study-of-multimodal-model","slug":"an-empirical-study-of-multimodal-model","title":"An Empirical Study of Multimodal Model Merging","date":"2023-04-28","arxiv_id":"2304.14933","repositories_listed":1,"syntology":{"n":1,"n_ran":1,"n_constructed":0,"n_ran_checked":0,"n_instrument":1,"n_unverified":0,"n_honours":0,"n_violates":0,"n_no_contract":0,"n_pointer_only":1,"phrase":"1 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 1 where Syntology's instrument failed) · 0 unverified","sample_list":"/paper/an-empirical-study-of-multimodal-model#ran","syntology_url":"https://syntology.ai/paper/2304.14933","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2304.14933"}},"official":{"repos":["ylsung/vl-merging"],"state":"official: no sample here; runs from other or unrecorded repositories","n_ran":0,"n_constructed":0,"n_ran_no_instrument_failure":0,"n_unverified":0,"ran_from_kinds":["unlocated"]}}},{"url":null,"slug":"transferring-visual-explainability-of-self","title":"Transferring Visual Explainability of Self-Explaining Models through Task Arithmetic","date":"2025-07-06","arxiv_id":"2507.04380","repositories_listed":0,"syntology":null},{"url":null,"slug":"duet-dual-incremental-object-detection-via","title":"DuET: Dual Incremental Object Detection via Exemplar-Free Task Arithmetic","date":"2025-06-26","arxiv_id":"2506.21260","repositories_listed":0,"syntology":null},{"url":null,"slug":"culturemert-continual-pre-training-for-cross","title":"CultureMERT: Continual Pre-Training for Cross-Cultural Music Representation Learning","date":"2025-06-21","arxiv_id":"2506.17818","repositories_listed":0,"syntology":null},{"url":null,"slug":"subspace-boosted-model-merging","title":"Subspace-Boosted Model Merging","date":"2025-06-19","arxiv_id":"2506.16506","repositories_listed":0,"syntology":null},{"url":null,"slug":"fedrpca-enhancing-federated-lora-aggregation","title":"FedRPCA: Enhancing Federated LoRA Aggregation Using Robust PCA","date":"2025-06-01","arxiv_id":"2506.01194","repositories_listed":0,"syntology":null},{"url":null,"slug":"on-fairness-of-task-arithmetic-the-role-of","title":"On Fairness of Task Arithmetic: The Role of Task Vectors","date":"2025-05-30","arxiv_id":"2505.24262","repositories_listed":0,"syntology":null},{"url":null,"slug":"scalable-strategies-for-continual-learning","title":"Scalable Strategies for Continual Learning with Replay","date":"2025-05-18","arxiv_id":"2505.12512","repositories_listed":0,"syntology":null},{"url":null,"slug":"2505-10859","title":"MCU: Improving Machine Unlearning through Mode Connectivity","date":"2025-05-16","arxiv_id":"2505.10859","repositories_listed":0,"syntology":null},{"url":null,"slug":"cat-merging-a-training-free-approach-for","title":"CAT Merging: A Training-Free Approach for Resolving Conflicts in Model Merging","date":"2025-05-11","arxiv_id":"2505.06977","repositories_listed":0,"syntology":null},{"url":null,"slug":"single-input-multi-output-model-merging","title":"Single-Input Multi-Output Model Merging: Leveraging Foundation Models for Dense Multi-Task Learning","date":"2025-04-15","arxiv_id":"2504.11268","repositories_listed":0,"syntology":null},{"url":null,"slug":"when-is-task-vector-provably-effective-for","title":"When is Task Vector Provably Effective for Model Editing? A Generalization Analysis of Nonlinear Transformers","date":"2025-04-15","arxiv_id":"2504.10957","repositories_listed":0,"syntology":null},{"url":null,"slug":"openthaigpt-1-6-and-r1-thai-centric-open","title":"OpenThaiGPT 1.6 and R1: Thai-Centric Open Source and Reasoning Large Language Models","date":"2025-04-02","arxiv_id":"2504.01789","repositories_listed":0,"syntology":null},{"url":null,"slug":"disentangling-task-interference-within","title":"Disentangling Task Interference within Neurons: Model Merging in Alignment with Neuronal Mechanisms","date":"2025-03-07","arxiv_id":"2503.05320","repositories_listed":0,"syntology":null},{"url":null,"slug":"layer-aware-task-arithmetic-disentangling","title":"Layer-Aware Task Arithmetic: Disentangling Task-Specific and Instruction-Following Knowledge","date":"2025-02-27","arxiv_id":"2502.20186","repositories_listed":0,"syntology":null},{"url":null,"slug":"neural-networks-remember-more-the-power-of","title":"Neural Networks Remember More: The Power of Parameter Isolation and Combination","date":"2025-02-16","arxiv_id":"2502.10966","repositories_listed":0,"syntology":null},{"url":null,"slug":"mediator-memory-efficient-llm-merging-with","title":"Mediator: Memory-efficient LLM Merging with Less Parameter Conflicts and Uncertainty Based Routing","date":"2025-02-06","arxiv_id":"2502.04411","repositories_listed":0,"syntology":null},{"url":null,"slug":"task-arithmetic-in-trust-region-a-training","title":"Task Arithmetic in Trust Region: A Training-Free Model Merging Approach to Navigate Knowledge Conflicts","date":"2025-01-25","arxiv_id":"2501.15065","repositories_listed":0,"syntology":null},{"url":null,"slug":"soup-to-go-mitigating-forgetting-during","title":"Soup to go: mitigating forgetting during continual learning with model averaging","date":"2025-01-09","arxiv_id":"2501.05559","repositories_listed":0,"syntology":null},{"url":null,"slug":"badtv-unveiling-backdoor-threats-in-third","title":"BADTV: Unveiling Backdoor Threats in Third-Party Task Vectors","date":"2025-01-04","arxiv_id":"2501.02373","repositories_listed":0,"syntology":null},{"url":null,"slug":"when-domain-generalization-meets-generalized","title":"When Domain Generalization meets Generalized Category Discovery: An Adaptive Task-Arithmetic Driven Approach","date":"2025-01-01","arxiv_id":null,"repositories_listed":0,"syntology":null},{"url":null,"slug":"bias-vector-mitigating-biases-in-language","title":"Bias Vector: Mitigating Biases in Language Models with Task Arithmetic Approach","date":"2024-12-16","arxiv_id":"2412.11679","repositories_listed":0,"syntology":null},{"url":null,"slug":"multi-task-model-merging-via-adaptive-weight","title":"Multi-Task Model Merging via Adaptive Weight Disentanglement","date":"2024-11-27","arxiv_id":"2411.18729","repositories_listed":0,"syntology":null},{"url":null,"slug":"task-arithmetic-through-the-lens-of-one-shot","title":"Task Arithmetic Through The Lens Of One-Shot Federated Learning","date":"2024-11-27","arxiv_id":"2411.18607","repositories_listed":0,"syntology":null},{"url":null,"slug":"beyond-task-vectors-selective-task-arithmetic","title":"Beyond Task Vectors: Selective Task Arithmetic Based on Importance Metrics","date":"2024-11-25","arxiv_id":"2411.16139","repositories_listed":0,"syntology":null},{"url":null,"slug":"atm-improving-model-merging-by-alternating","title":"ATM: Improving Model Merging by Alternating Tuning and Merging","date":"2024-11-05","arxiv_id":"2411.03055","repositories_listed":0,"syntology":null},{"url":null,"slug":"efficient-and-effective-weight-ensembling","title":"Efficient and Effective Weight-Ensembling Mixture of Experts for Multi-Task Model Merging","date":"2024-10-29","arxiv_id":"2410.21804","repositories_listed":0,"syntology":null},{"url":null,"slug":"the-non-local-model-merging-problem","title":"The Non-Local Model Merging Problem: Permutation Symmetries and Variance Collapse","date":"2024-10-16","arxiv_id":"2410.12766","repositories_listed":0,"syntology":null},{"url":null,"slug":"what-matters-for-model-merging-at-scale","title":"What Matters for Model Merging at Scale?","date":"2024-10-04","arxiv_id":"2410.03617","repositories_listed":0,"syntology":null},{"url":null,"slug":"task-arithmetic-for-language-expansion-in","title":"Task Arithmetic for Language Expansion in Speech Translation","date":"2024-09-17","arxiv_id":"2409.11274","repositories_listed":0,"syntology":null},{"url":null,"slug":"metagpt-merging-large-language-models-using","title":"MetaGPT: Merging Large Language Models Using Model Exclusive Task Arithmetic","date":"2024-06-17","arxiv_id":"2406.11385","repositories_listed":0,"syntology":null},{"url":null,"slug":"on-giant-s-shoulders-effortless-weak-to","title":"On Giant's Shoulders: Effortless Weak to Strong by Dynamic Logits Fusion","date":"2024-06-17","arxiv_id":"2406.15480","repositories_listed":0,"syntology":null},{"url":null,"slug":"syn2real-leveraging-task-arithmetic-for","title":"Task Arithmetic can Mitigate Synthetic-to-Real Gap in Automatic Speech Recognition","date":"2024-06-05","arxiv_id":"2406.02925","repositories_listed":0,"syntology":null},{"url":null,"slug":"hpe-cogvlm-new-head-pose-grounding-task","title":"HPE-CogVLM: Advancing Vision Language Models with a Head Pose Grounding Task","date":"2024-06-04","arxiv_id":"2406.01914","repositories_listed":0,"syntology":null},{"url":null,"slug":"to-each-textual-sequence-its-own-improving","title":"To Each (Textual Sequence) Its Own: Improving Memorized-Data Unlearning in Large Language Models","date":"2024-05-06","arxiv_id":"2405.03097","repositories_listed":0,"syntology":null},{"url":null,"slug":"ethos-rectifying-language-models-in","title":"Ethos: Rectifying Language Models in Orthogonal Parameter Space","date":"2024-03-13","arxiv_id":"2403.08994","repositories_listed":0,"syntology":null},{"url":null,"slug":"language-and-task-arithmetic-with-parameter","title":"Language and Task Arithmetic with Parameter-Efficient Layers for Zero-Shot Summarization","date":"2023-11-15","arxiv_id":"2311.09344","repositories_listed":0,"syntology":null},{"url":null,"slug":"task-arithmetic-with-lora-for-continual","title":"Task Arithmetic with LoRA for Continual Learning","date":"2023-11-04","arxiv_id":"2311.02428","repositories_listed":0,"syntology":null}],"syntology_records":18,"syntology_note":"a paper without a record is not a recorded non-run: it may lack an arXiv id or simply be absent from the graph layer"},"key_notes":{"n_ran_checked":"legacy name, kept unchanged so existing readers do not break: it counts the samples that ran with no instrument failure (honoured, violated, and ran with no contract checked); it does not mean a contract was checked, and the pages print it as 'K with no instrument failure', not 'K checked'","n_constructed":"a sub-count of the samples that ran, never subtracted from them and never a failure: an executed sample whose run returned an instance of its own class (fixture_out_type equals the entry name): the run built an object and did not compute a result (Syntology's RAN record, counts.constructed)"},"since_archive_judged":{"papers":0,"papers_placed":0,"judged_by":{"model":"a model","models_as_recorded":[],"from":"lb_rows_run_verified.jsonl: each record's `model` field (the extractor's call that wrote decisions[])"},"note":"a model's judgement from the paper's own tables that they report on this leaderboard, not a result: no value from a paper that was not placed is printed or carried here; papers counts each paper once over these leaderboards","backfill":"8,886 of 18,637 newer papers checked so far (733 more had nothing to place; 41 could not be read)","backfill_complete":false,"leaderboards":[]},"papers_where_code_ran":18,"papers_where_code_ran_first_page":"/task/task-arithmetic/papers/ran/1","papers_where_code_ran_caption":"We ran code from the paper's repository; we did not run it on this task or check it against the task's benchmarks.","pages":[{"order":"ran","page":1,"of":1,"url":"/task/task-arithmetic/papers/ran/1","json":"/task/task-arithmetic/papers/ran/1.json","rows":18}],"description_links":{"kept":0,"unwrapped_to_text":0,"bare_urls_linked":0,"relative_images_dropped":0,"rule":"internal links are kept only when the target slug exists in the catalog"},"syntology":{"read_at":"2026-10-05T00:36:11+00:00","claim":"Per-sample execution status on synthesized fixtures ('ran N of M samples'); not a correctness claim and not a ranking signal.","status_vocabulary":{"ran_honours":"ran, honoured the contract we drafted","ran_violates":"ran, violated the contract we drafted","ran_draft_wrong":"ran; our contract draft was wrong, not the code","ran_fixture":"ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)","ran":"ran on a synthesized input","unverified":"unverified (harvested, no recorded run)"},"row_route":"each syntology object on a benchmark's first row and on a paper carries sample_list (this site's paper page, #ran), syntology_url (syntology.ai's page for the paper, only when that route probed live) and mcp (the agent call that lists the paper's harvested samples; not a verdict on the paper)","developers":"https://syntology.ai/developers"},"not_shown":{"libraries":"the archive has no per-task library table","trend_sparklines":"the Trend column of the benchmarks table was a rendered image; it is not in the archive","social_and_latest_sorts":"stars and social signals are not in the archive"},"record_sha256":"643d6e936ca3148472a89efad1b27039678a48138d03b218b36009ec791c431b","record_changed_at":"2026-10-05","record_changed_at_basis":"changed"}