Mixture-Trained Merging (MTM) is a method for creating unified language models that combine multiple objectives—such as mathematics, code, instruction following, and controllable thinking—into a single parameter set. Instead of sequentially post‑training on each objective, MTM trains each branch on a mixture of objectives, ensuring that the branches remain compatible in weight space and can be merged without degrading performance. The approach iteratively refines merge coefficients using low‑cost evaluations and multi‑objective Bayesian optimization, outperforming naive merging and preserving distinct behaviors across domains.
By SeongHyeon Kim, Chaeyun Jang, Seungyoo Lee, Jiyeon Ham, Yunju Bak, Boseop Kim, Juho Lee
arXiv:2606. 05613v1 Announce Type: new Abstract: The rapid evolution of Large Language Models (LLMs) has established cross-lingual versatility as a defining feature of modern systems.
By Long P. Hoang, Yiran Zhao, Wei Lu, Wenxuan Zhang
arXiv:2510. 01167v2 Announce Type: replace-cross Abstract: Aligning large language models to human preferences is inherently multidimensional, yet most pipelines collapse heterogeneous signals into a single objective.
By Yiran Shen, Yu Xia, Jonathan Chang, Prithviraj Ammanabrolu
arXiv:2610.23062v1 Announce Type: new
Abstract: Multi-objective optimization models are often deployed in settings where decision makers (DMs) must balance some conflicting criteria. In practice, how...
By Gokhan Ceyhan
The paper critiques current multi‑task active learning (MTL‑AL) methods that use linear scalarization of objectives like uncertainty, diversity, and gradient conflict, arguing that such combinations miss Pareto‑optimal trade‑offs. It introduces a new framework that separates acquisition into informativeness and exploration, decouples conflict scoring from candidate ranking, and uses gradient surgery during training. The authors also propose a cross‑task mutual information metric and demonstrate consistent improvements on COCO, Cityscapes, and NYUv2 datasets compared to recent baselines.
By Juho Jung, Seongjae Kang, Jinyoung Han
The paper introduces DualSFT, a one-shot algorithm that simultaneously selects a parameter mask and a data subset for fine‑tuning large language models by using a shared gradient‑interaction matrix. By casting both selections as bilevel problems under a common validation objective, the authors derive a unified scoring rule that aggregates parameter importance and data utility as column‑wise and row‑wise sums of the same matrix. Experiments on 3B‑9B LLMs show that DualSFT improves target‑task performance and stability‑plasticity trade‑offs compared to single‑axis variants and outperforms sequential hybrid baselines under matched budgets.
By Xinrui Chen, Liu Yang, Ou Wu
MOSCOPT is a parameter‑free algorithm that jointly optimizes a pool of N skills and a gating skill G, which dynamically selects K skills at each step for LLM agents. It introduces EditAdam, a dual‑state optimizer, and a three‑phase interleaved update scheme that monotonically improves performance without gradient or parameter tuning. Experiments on five benchmarks and three target LLMs show MOSCOPT consistently outperforms baselines, highlighting the importance of mixture‑of‑skills architecture and collective evolution.
By Zhenyu Zhang1, Jiudong Yang
arXiv:2609.08572v1 Announce Type: new
Abstract: Large language model (LLM)-based multi-agent systems (MAS) achieve strong performance by employing specialized multiple agents, yet their performance d...
By Jaewon Chu, Jinwoo Seo, Jaewon Cho, Jeehye Na, Yunyang Xiong, Youngdae Kim, Hyunwoo J. Kim
arXiv:2607. 25675v1 Announce Type: new Abstract: Text-space optimization adapts large language models (LLMs) by editing external natural-language artifacts rather than model weights, so the optimized artifacts remain inspectable and the model can be treated as a black box.
By Jiangwang Chen, Zixin Song, Junlin Liu, Shuaiyu Zhou, Haiyan Wu, Haihan Shi, Chenxi Zhou, Hanqing Li, Xiao Yang, Da Zhu, Guanjun Jiang, Hai Wan, Xibin Zhao
The paper introduces EoupCT, a framework that estimates and orthogonalizes unknown pre‑training gradients to mitigate catastrophic forgetting during continual fine‑tuning of large language models. It generates pseudo data most susceptible to forgetting using a learnable soft prompt with Gumbel‑Softmax, then jointly optimizes model parameters and the prompt via a first‑order Pareto optimizer to enforce orthogonality between new task updates and the estimated gradients. Experiments on multiple LLMs show that EoupCT preserves both task‑specific performance and the models’ inherent general‑purpose knowledge.
By Bing Wang, Changchun Li, Xin-Qiang Cai, Lin Yuanbo Wu, Ximing Li, Gang Niu, Masashi Sugiyama
arXiv:2609.23716v1 Announce Type: cross
Abstract: Textual-gradient methods automate prompt optimization through natural-language feedback, but their iterative updates can be unstable. We identify two...
By Yifan Xu, Yixuan Li, Xinzhuo Li, Yixin Gu, Yifan Shen, Lijun Yu, Haohan Wang
CoMerge is a conflict‑driven preference optimization framework for merging multiple expert language models into a single multi‑task model without full retraining. It treats model merging as a preference optimization problem, using self‑supervised, conflict‑driven hard negative samples derived from naive merging defects to refine lightweight, tensor‑wise merging coefficients. Experiments show CoMerge achieves near‑perfect performance on MergeBench and improves instruction‑following and safety on Llama‑3.1‑8B‑Instruct while optimizing only 1,445 scalar coefficients.
By Mingjie Zheng, Zihao Chen, Wenqing Chen, Weile Yuan, Zhixuan Chu, Jianxing Yu, Zibin Zheng