Architecture Decoupling Is Not All You Need For Unified Multimodal Model

• URL: http://arxiv.org/abs/2511.22663v1
• Date: Thu, 27 Nov 2025 17:55:25 GMT
• Title: Architecture Decoupling Is Not All You Need For Unified Multimodal Model
• Authors: Dian Zheng, Manyuan Zhang, Hongyu Li, Kai Zou, Hongbo Liu, Ziyu Guo, Kaituo Feng, Yexin Liu, Ying Luo, Yan Feng, Peng Pei, Xunliang Cai, Hongsheng Li,
• Abstract summary: We propose Attention Interaction Alignment (AIA) loss, which explicitly learns Task-Specific multimodal interaction patterns during training.<n>AIA not only refines cross-modal attention patterns, but also boosts both generation and understanding performance.
• Score: 64.19284951218098
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Unified multimodal models for image generation and understanding represent a significant step toward AGI and have attracted widespread attention from researchers. The main challenge of this task lies in the difficulty in establishing an optimal training paradigm due to inherent conflicting targets in understanding and generation tasks. To alleviate these conflicts and pursue higher performance, many researchers adopt varying degrees of model decoupling (e.g., Double image encoders, MOE/MOT architecture, or frozen MLLM). However, excessive model decoupling can lead to the loss of interleave generation ability, undermining the original intent of unified models. In this work, we aim to explore how to mitigate task conflicts without resorting to model decoupling. Firstly, we analyze why decoupling alleviates conflicts by studying the cross-modal attention behavior of models. We observe that model decoupling essentially drives models toward task-specific multimodal interaction patterns, as seen in Qwen-VL and HunyuanImage, and that the more thorough the decoupling, the more consistent the behavior becomes. Motivated by this observation, we propose Attention Interaction Alignment (AIA) loss, which explicitly learns Task-Specific multimodal interaction patterns during training. To demonstrate the generalizability of our AIA loss, we apply it to Emu3 and Janus-Pro during SFT and post-training stage respectively. Without bells and whistles, AIA not only refines cross-modal attention patterns, but also boosts both generation and understanding performance.

Related papers

• Crab$^{+}$: A Scalable and Unified Audio-Visual Scene Understanding Model with Explicit Cooperation [83.75249714794977]
  We present Crab$+$, a scalable and unified audio-visual scene understanding model.<n>On the data side, we introduce AV-UIE v2, a comprehensive Audio-Visual Unified Instruction-tuning dataset.<n>On the model side, we design a unified interface to align heterogeneous task formulations.<n>We successfully reverse the negative transfer trend, achieving positive transfer where multi-task learning surpasses single-task baselines in nearly 88% of tasks.
  arXiv  Detail & Related papers  (2026-03-04T14:43:57Z)
• UniG2U-Bench: Do Unified Models Advance Multimodal Understanding? [50.92401586025528]
  Unified multimodal models have recently demonstrated strong generative capabilities, yet whether and when generation improves understanding remains unclear.<n>We introduce UniG2U-Bench, a comprehensive benchmark categorizing generation-to-understanding (G2U) evaluation into 7 regimes and 30 subtasks.
  arXiv  Detail & Related papers  (2026-03-03T18:36:16Z)
• Model Merging in the Essential Subspace [78.5390284258307]
  Model merging aims to integrate multiple task-specific fine-tuned models into a single multi-task model without additional training.<n>Despite extensive research, task interference remains a major obstacle that often undermines the performance of merged models.<n>We propose ESM (Essential Subspace Merging), a robust framework for effective model merging.
  arXiv  Detail & Related papers  (2026-02-23T00:33:38Z)
• Mind the Gap: Aligning Vision Foundation Models to Image Feature Matching [31.42132290162457]
  We introduce a new framework called IMD (Image feature Matching with a pre-trained Diffusion model) with two parts.<n>Unlike the dominant solutions employing contrastive-learning based foundation models that emphasize global semantics, we integrate the generative-based diffusion models.<n>Our proposed IMD establishes a new state-of-the-art in commonly evaluated benchmarks, and the superior 12% improvement in IMIM indicates our method efficiently mitigates the misalignment.
  arXiv  Detail & Related papers  (2025-07-14T14:28:15Z)
• UniFork: Exploring Modality Alignment for Unified Multimodal Understanding and Generation [39.921363034430875]
  Unified image understanding and generation has emerged as a promising paradigm in multimodal artificial intelligence.<n>We study the modality alignment behaviors of task-specific expert models for understanding and generation.<n>We introduce UniFork, a novel Y-shaped architecture that shares the shallow layers for cross-task representation learning, while employing task-specific branches in deeper layers to avoid task interference.
  arXiv  Detail & Related papers  (2025-06-20T17:52:31Z)
• Resolving Task Objective Conflicts in Unified Model via Task-Aware Mixture-of-Experts [11.790264535536965]
  multimodal large language models (MLLMs) integrate both understanding and generation tasks within a single framework.<n> intrinsic Task Objective Conflicts between high-level semantic abstraction in understanding and fine-grained detail preservation in generation pose significant challenges.<n>We propose a novel approach that decouples internal components of AR to resolve task objective conflicts.
  arXiv  Detail & Related papers  (2025-06-04T05:44:21Z)
• Modeling Multi-Task Model Merging as Adaptive Projective Gradient Descent [72.10987117380584]
  Merging multiple expert models offers a promising approach for performing multi-task learning without accessing their original data.<n>We find existing methods discard task-specific information that, while causing conflicts, is crucial for performance.<n>Our approach consistently outperforms previous methods, achieving state-of-the-art results across diverse architectures and tasks in both vision and NLP domains.
  arXiv  Detail & Related papers  (2025-01-02T12:45:21Z)
• Concrete Subspace Learning based Interference Elimination for Multi-task Model Fusion [86.6191592951269]
  Merging models fine-tuned from common extensively pretrained large model but specialized for different tasks has been demonstrated as a cheap and scalable strategy to construct a multitask model that performs well across diverse tasks. We propose the CONtinuous relaxation dis (Concrete) subspace learning method to identify a common lowdimensional subspace and utilize its shared information track interference problem without sacrificing performance.
  arXiv  Detail & Related papers  (2023-12-11T07:24:54Z)
• AdaMerging: Adaptive Model Merging for Multi-Task Learning [68.75885518081357]
  This paper introduces an innovative technique called Adaptive Model Merging (AdaMerging) It aims to autonomously learn the coefficients for model merging, either in a task-wise or layer-wise manner, without relying on the original training data. Compared to the current state-of-the-art task arithmetic merging scheme, AdaMerging showcases a remarkable 11% improvement in performance.
  arXiv  Detail & Related papers  (2023-10-04T04:26:33Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.