{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,10]],"date-time":"2026-08-10T19:37:25Z","timestamp":1786390645693,"version":"build-2736575974"},"reference-count":47,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"9","license":[{"start":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T00:00:00Z","timestamp":1788220800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T00:00:00Z","timestamp":1788220800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,9,1]],"date-time":"2026-09-01T00:00:00Z","timestamp":1788220800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"name":"Major Research Plan of the National Natural Science Foundation of China","award":["92370110"],"award-info":[{"award-number":["92370110"]}]},{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"crossref","award":["62576038"],"award-info":[{"award-number":["62576038"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. Knowl. Data Eng."],"published-print":{"date-parts":[[2026,9]]},"DOI":"10.1109\/tkde.2026.3708509","type":"journal-article","created":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T21:44:30Z","timestamp":1782855870000},"page":"5867-5878","source":"Crossref","is-referenced-by-count":0,"title":["ExMe: Keep Improving With Extrapolation and Merging"],"prefix":"10.1109","volume":"38","author":[{"ORCID":"https:\/\/orcid.org\/0009-0002-7837-2317","authenticated-orcid":false,"given":"Yiguan","family":"Lin","sequence":"first","affiliation":[{"name":"Beijing Institute of Technology","place":["Beijing, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Bin","family":"Xu","sequence":"additional","affiliation":[{"name":"Beijing Institute of Technology","place":["Beijing, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yinghao","family":"Li","sequence":"additional","affiliation":[{"name":"Shanxi University","place":["Taiyuan, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2422-0548","authenticated-orcid":false,"given":"Yang","family":"Gao","sequence":"additional","affiliation":[{"name":"Beijing Institute of Technology","place":["Beijing, China"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1007\/s11704-026-60308-3"},{"key":"ref2","first-page":"608","article-title":"Retrieval-augmented generation for social media content creation with sentence window and auto-merging retrieval","volume-title":"Proc. Int. Conf. Informat., Multimedia, Cyber Inf. Syst","author":"Prawira"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2025.3542036"},{"key":"ref4","doi-asserted-by":"publisher","DOI":"10.1109\/ICCE63647.2025.10929957"},{"key":"ref5","article-title":"Finetuned language models are zero-shot learners","volume-title":"Proc. 10th Int. Conf. Learn. Representations","author":"Wei"},{"key":"ref6","doi-asserted-by":"publisher","DOI":"10.52202\/068431-2011"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.52202\/068431-2176"},{"key":"ref8","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.findings-emnlp.1012"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.acl-demos.9"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.emnlp-main.632"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP49660.2025.10890406"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP49660.2025.10890428"},{"key":"ref13","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP49660.2025.10888591"},{"key":"ref14","article-title":"Dataless knowledge fusion by merging weights of language models","volume-title":"Proc. 11th Int. Conf. Learn. Representations","author":"Jin"},{"key":"ref15","article-title":"Editing models with task arithmetic","volume-title":"Proc. 11th Int. Conf. Learn. Representations","author":"Ilharco"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.52202\/075280-0310"},{"key":"ref17","first-page":"57755","article-title":"Language models are super Mario: Absorbing abilities from homologous models as a free lunch","volume-title":"Proc. 41st Int. Conf. Mach. Learn.","volume":"235","author":"Yu","year":"2024"},{"key":"ref18","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.emnlp-main.35"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2025.acl-long.51"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.52202\/075280-2338"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2021.acllong.353"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2021.emnlp-main.243"},{"key":"ref23","article-title":"LoRa: Low-rank adaptation of large language models","volume-title":"Proc. 10th Int. Conf. Learn. Representations","author":"Hu"},{"key":"ref24","first-page":"23965","article-title":"Model soups: Averaging weights of multiple fine-tuned models improves accuracy without increasing inference time","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Wortsman"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.52202\/068431-1287"},{"key":"ref26","first-page":"39713","article-title":"Maximizing intermediate checkpoint value in LLM pretraining with Bayesian optimization","volume-title":"Proc. 42nd Int. Conf. Mach. Learn.","volume":"267","author":"Liu","year":"2025"},{"key":"ref27","article-title":"Mixture of LoRa experts","volume-title":"Proc. 12th Int. Conf. Learn. Representations","author":"Wu"},{"key":"ref28","doi-asserted-by":"publisher","DOI":"10.52202\/079017-2691"},{"key":"ref29","article-title":"Model merging with SVD to tie the knots","volume-title":"Proc. 13th Int. Conf. Learn. Representations","author":"Stoica"},{"key":"ref30","first-page":"1309","article-title":"Essentially no barriers in neural network energy landscape","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Draxler"},{"key":"ref31","article-title":"The role of permutation invariance in linear mode connectivity of neural networks","author":"Entezari","year":"2021"},{"key":"ref32","article-title":"Loss surfaces, mode connectivity, and fast ensembling of DNNs","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Garipov"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.52202\/075280-2913"},{"key":"ref34","article-title":"Fine-tuning attention modules only: Enhancing weight disentanglement in task arithmetic","author":"Jin","year":"2024"},{"key":"ref35","first-page":"11615","article-title":"Uniform convergence may be unable to explain generalization in deep learning","volume-title":"Proc. Adv. Neural Inf. Process. Syst.","author":"Nagarajan"},{"key":"ref36","article-title":"Qwen3 technical report","author":"Yang","year":"2025"},{"key":"ref37","article-title":"Instruction tuning with GPT-4","author":"Peng","year":"2023"},{"key":"ref38","article-title":"OpenCompass: A universal evaluation platform for foundation models","author":"Contributors","year":"2023"},{"key":"ref39","article-title":"Training verifiers to solve math word problems.","author":"Cobbe","year":"2021"},{"key":"ref40","article-title":"Measuring mathematical problem solving with the MATH dataset","volume-title":"Proc. Neural Inf. Process. Syst. Track Datasets Benchmarks 1","author":"Hendrycks"},{"key":"ref41","article-title":"Evaluating large language models trained on code.","author":"Chen","year":"2021"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.findings-acl.671"},{"key":"ref43","doi-asserted-by":"publisher","DOI":"10.52202\/075280-2749"},{"key":"ref44","article-title":"Evaluating the performance of large language models on GAOKAO benchmark","author":"Zhang","year":"2023"},{"key":"ref45","article-title":"American invitational mathematics examination (AIME) 2024","author":"Zhang","year":"2024"},{"key":"ref46","doi-asserted-by":"publisher","DOI":"10.2307\/jj.19985101.13"},{"key":"ref47","article-title":"The Llama 3 herd of models","author":"Grattafiori","year":"2024"}],"container-title":["IEEE Transactions on Knowledge and Data Engineering"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/69\/11644801\/11589567.pdf?arnumber=11589567","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,10]],"date-time":"2026-08-10T19:04:52Z","timestamp":1786388692000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/11589567\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9]]},"references-count":47,"journal-issue":{"issue":"9"},"URL":"https:\/\/doi.org\/10.1109\/tkde.2026.3708509","relation":{},"ISSN":["1041-4347","1558-2191","2326-3865"],"issn-type":[{"value":"1041-4347","type":"print"},{"value":"1558-2191","type":"electronic"},{"value":"2326-3865","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9]]}}}