The Flow has not summarised this story yet — read it at Hugging Face Trending Papers.
The paper introduces a new multi‑task model‑merging framework that tackles task interference by projecting task vectors into a high‑dimensional sparse feature space using Sparse Autoencoders, enabling feature‑level disentanglement before fusion. It also proposes a lightweight Group‑Ranked Zeroth‑Order Optimizer to identify task‑critical layers for selective merging, reducing computational overhead. Experiments on Qwen2.5‑1.5B and Qwen2.5‑7B show consistent performance gains over several baselines across reasoning, code generation, instruction following, and general knowledge tasks, with a 2.78% improvement in a highly conflicting four‑task setting.
arXiv:2312. 06173v2 Announce Type: replace Abstract: Merging models fine-tuned from a common, extensively pre-trained large model but specialized for different tasks has been demonstrated as a cheap and scalable strategy to construct a multi-task model that performs well across diverse tasks.
arXiv:2606. 18627v1 Announce Type: new Abstract: Model merging has emerged as a training-free alternative to multi-task learning, aiming to combine multiple task-specific fine-tuned models into a single multi-task model.
arXiv:2606. 19164v1 Announce Type: cross Abstract: Model merging aims to enable multi-task learning by integrating the capabilities of multiple models fine-tuned from the same pre-trained checkpoint into a single model.
The paper introduces Resolving Interference (RI), a lightweight adaptation framework that reduces cross‑task interference when merging expert models. RI disentangles expert models to be functionally orthogonal to other tasks, using only unlabeled auxiliary data, and improves merging performance by up to 3.8% and generalization to unseen domains by up to 2.3%. The method is robust to the source of auxiliary input and less sensitive to hyperparameter tuning.
CRLoMTL is a parameter‑efficient multi‑task learning framework that mitigates task interference by operating in a low‑rank parameter‑update space rather than gradient space. It learns task‑specific LoRA updates and fuses them through a shared‑subspace resolver based on singular value decomposition, producing a single deployment‑time update that can be folded into the backbone. Experiments on Swin Transformer backbones across PASCAL‑Context, NYUD‑v2, and CelebA show that CRLoMTL improves accuracy over single‑task baselines while training only a small fraction of the parameters.