LLaVA-CMoE: Towards Continual Mixture of Experts for Large Vision-Language Models

• URL: http://arxiv.org/abs/2503.21227v3
• Date: Wed, 25 Jun 2025 08:30:20 GMT
• Title: LLaVA-CMoE: Towards Continual Mixture of Experts for Large Vision-Language Models
• Authors: Hengyuan Zhao, Ziqin Wang, Qixin Sun, Kaiyou Song, Yilin Li, Xiaolin Hu, Qingpei Guo, Si Liu,
• Abstract summary: LLaVA-CMoE is a continual learning framework for large language models.<n> Probe-Guided Knowledge Extension mechanism determines when and where new experts should be added.<n>Probabilistic Task Locator assigns each task a dedicated, lightweight router.
• Score: 21.888139819188105
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Mixture of Experts (MoE) architectures have recently advanced the scalability and adaptability of large language models (LLMs) for continual multimodal learning. However, efficiently extending these models to accommodate sequential tasks remains challenging. As new tasks arrive, naive model expansion leads to rapid parameter growth, while modifying shared routing components often causes catastrophic forgetting, undermining previously learned knowledge. To address these issues, we propose LLaVA-CMoE, a continual learning framework for LLMs that requires no replay data of previous tasks and ensures both parameter efficiency and robust knowledge retention. Our approach introduces a Probe-Guided Knowledge Extension mechanism, which uses probe experts to dynamically determine when and where new experts should be added, enabling adaptive and minimal parameter expansion tailored to task complexity. Furthermore, we present a Probabilistic Task Locator that assigns each task a dedicated, lightweight router. To handle the practical issue that task labels are unknown during inference, we leverage a VAE-based reconstruction strategy to identify the most suitable router by matching input distributions, allowing automatic and accurate expert allocation. This design mitigates routing conflicts and catastrophic forgetting, enabling robust continual learning without explicit task labels. Extensive experiments on the CoIN benchmark, covering eight diverse VQA tasks, demonstrate that LLaVA-CMoE delivers strong continual learning performance with a compact model size, significantly reducing forgetting and parameter overhead compared to prior methods. These results showcase the effectiveness and scalability of our approach for parameter-efficient continual learning in large language models. Our code will be open-sourced soon.

Related papers

• Dynamic Mixture of Progressive Parameter-Efficient Expert Library for Lifelong Robot Learning [69.81148368677593]
  A generalist agent must continuously learn and adapt throughout its lifetime, achieving efficient forward transfer while minimizing catastrophic forgetting.<n>Previous work has explored parameter-efficient fine-tuning for single-task adaptation, effectively steering a frozen pretrained model with a small number of parameters.<n>We propose Dynamic Mixture of Progressive Efficient Expert Library (DMPEL) for lifelong robot learning.<n>Our framework outperforms state-of-the-art lifelong learning methods in success rates across continual adaptation, while utilizing minimal trainable parameters and storage.
  arXiv  Detail & Related papers  (2025-06-06T11:13:04Z)
• Route-and-Reason: Scaling Large Language Model Reasoning with Reinforced Model Router [9.580226379350737]
  Multi-step reasoning has proven essential for enhancing the problem-solving capabilities of Large Language Models.<n>Yet, many reasoning steps are relatively simple and can be handled by more efficient smaller-scale language models.<n>We propose R2-Reasoner, a novel framework that enables collaborative reasoning across heterogeneous LLMs.
  arXiv  Detail & Related papers  (2025-06-06T09:18:56Z)
• MoTE: Mixture of Task-specific Experts for Pre-Trained ModelBased Class-incremental Learning [39.892628170627496]
  Class-incremental learning (CIL) requires deep learning models to continuously acquire new knowledge from streaming data.<n> prompt-based approaches suffer from prompt overwriting, while adapter-based methods face challenges such as dimensional misalignment between tasks.<n>We propose a mixture of task-specific experts (MoTE) framework that effectively mitigates the miscalibration caused by inconsistent output dimensions.
  arXiv  Detail & Related papers  (2025-05-21T03:06:10Z)
• FM-LoRA: Factorized Low-Rank Meta-Prompting for Continual Learning [19.068489119024388]
  Continual learning has emerged as a promising approach to leverage pre-trained models for sequential tasks.<n>Many existing CL methods incrementally store additional learned structures, such as Low-Rank Adaptation (LoRA) adapters or prompts.<n>We propose FM-LoRA, a novel and efficient low-rank adaptation method that integrates both a dynamic rank selector (DRS) and dynamic meta-prompting (DMP)
  arXiv  Detail & Related papers  (2025-04-09T19:36:18Z)
• Analytic Subspace Routing: How Recursive Least Squares Works in Continual Learning of Large Language Model [6.42114585934114]
  Large Language Models (LLMs) possess capabilities that can process diverse language-related tasks.<n>Continual Learning in Large Language Models (LLMs) arises which aims to continually adapt the LLMs to new tasks.<n>This paper proposes Analytic Subspace Routing(ASR) to address these challenges.
  arXiv  Detail & Related papers  (2025-03-17T13:40:46Z)
• Fresh-CL: Feature Realignment through Experts on Hypersphere in Continual Learning [3.708456605408296]
  We propose a method called Feature Realignment through Experts on hyperSpHere in Continual Learning (Fresh-CL) We propose a dynamic extension of ETF through mixture of experts, enabling adaptive projections onto diverse subspaces to enhance feature representation. Experiments on 11 datasets demonstrate a 2% improvement in accuracy compared to the strongest baseline.
  arXiv  Detail & Related papers  (2025-01-04T05:20:53Z)
• Reference Trustable Decoding: A Training-Free Augmentation Paradigm for Large Language Models [79.41139393080736]
  Large language models (LLMs) have rapidly advanced and demonstrated impressive capabilities. In-Context Learning (ICL) and. Efficient Fine-Tuning (PEFT) are currently two mainstream methods for augmenting. LLMs to downstream tasks. We propose Reference Trustable Decoding (RTD), a paradigm that allows models to quickly adapt to new tasks without fine-tuning.
  arXiv  Detail & Related papers  (2024-09-30T10:48:20Z)
• MoExtend: Tuning New Experts for Modality and Task Extension [61.29100693866109]
  MoExtend is an effective framework designed to streamline the modality adaptation and extension of Mixture-of-Experts (MoE) models. MoExtend seamlessly integrates new experts into pre-trained MoE models, endowing them with novel knowledge without the need to tune pretrained models.
  arXiv  Detail & Related papers  (2024-08-07T02:28:37Z)
• Diversifying the Expert Knowledge for Task-Agnostic Pruning in Sparse Mixture-of-Experts [75.85448576746373]
  We propose a method of grouping and pruning similar experts to improve the model's parameter efficiency. We validate the effectiveness of our method by pruning three state-of-the-art MoE architectures. The evaluation shows that our method outperforms other model pruning methods on a range of natural language tasks.
  arXiv  Detail & Related papers  (2024-07-12T17:25:02Z)
• PECTP: Parameter-Efficient Cross-Task Prompts for Incremental Vision Transformer [76.39111896665585]
  Incremental Learning (IL) aims to learn deep models on sequential tasks continually. Recent vast pre-trained models (PTMs) have achieved outstanding performance by prompt technique in practical IL without the old samples.
  arXiv  Detail & Related papers  (2024-07-04T10:37:58Z)
• Multi-Task Dense Prediction via Mixture of Low-Rank Experts [35.11968315125389]
  We present a novel decoder-focused method for multi-task dense prediction, called Mixture-of-Low-Rank-Experts (MLoRE) To model the global task relationships, MLoRE adds a generic convolution path to the original MoE structure, where each task feature can go through this path for explicit parameter sharing. Our experiments show that our MLoRE achieves superior performance compared to previous state-of-the-art methods on all metrics.
  arXiv  Detail & Related papers  (2024-03-26T14:40:17Z)
• Towards Robust Continual Learning with Bayesian Adaptive Moment Regularization [51.34904967046097]
  Continual learning seeks to overcome the challenge of catastrophic forgetting, where a model forgets previously learnt information. We introduce a novel prior-based method that better constrains parameter growth, reducing catastrophic forgetting. Results show that BAdam achieves state-of-the-art performance for prior-based methods on challenging single-headed class-incremental experiments.
  arXiv  Detail & Related papers  (2023-09-15T17:10:51Z)
• Dense Network Expansion for Class Incremental Learning [61.00081795200547]
  State-of-the-art approaches use a dynamic architecture based on network expansion (NE), in which a task expert is added per task. A new NE method, dense network expansion (DNE), is proposed to achieve a better trade-off between accuracy and model complexity. It outperforms the previous SOTA methods by a margin of 4% in terms of accuracy, with similar or even smaller model scale.
  arXiv  Detail & Related papers  (2023-03-22T16:42:26Z)
• A Dirichlet Process Mixture of Robust Task Models for Scalable Lifelong Reinforcement Learning [11.076005074172516]
  reinforcement learning algorithms can easily encounter catastrophic forgetting or interference when faced with lifelong streaming information. We propose a scalable lifelong RL method that dynamically expands the network capacity to accommodate new knowledge. We show that our method successfully facilitates scalable lifelong RL and outperforms relevant existing methods.
  arXiv  Detail & Related papers  (2022-05-22T09:48:41Z)
• Combining Modular Skills in Multitask Learning [149.8001096811708]
  A modular design encourages neural models to disentangle and recombine different facets of knowledge to generalise more systematically to new tasks. In this work, we assume each task is associated with a subset of latent discrete skills from a (potentially small) inventory. We find that the modular design of a network significantly increases sample efficiency in reinforcement learning and few-shot generalisation in supervised learning.
  arXiv  Detail & Related papers  (2022-02-28T16:07:19Z)
• Efficient Feature Transformations for Discriminative and Generative Continual Learning [98.10425163678082]
  We propose a simple task-specific feature map transformation strategy for continual learning. Theses provide powerful flexibility for learning new tasks, achieved with minimal parameters added to the base architecture. We demonstrate the efficacy and efficiency of our method with an extensive set of experiments in discriminative (CIFAR-100 and ImageNet-1K) and generative sequences of tasks.
  arXiv  Detail & Related papers  (2021-03-25T01:48:14Z)
• Reparameterizing Convolutions for Incremental Multi-Task Learning without Task Interference [75.95287293847697]
  Two common challenges in developing multi-task models are often overlooked in literature. First, enabling the model to be inherently incremental, continuously incorporating information from new tasks without forgetting the previously learned ones (incremental learning) Second, eliminating adverse interactions amongst tasks, which has been shown to significantly degrade the single-task performance in a multi-task setup (task interference)
  arXiv  Detail & Related papers  (2020-07-24T14:44:46Z)
• Meta Reinforcement Learning with Autonomous Inference of Subtask Dependencies [57.27944046925876]
  We propose and address a novel few-shot RL problem, where a task is characterized by a subtask graph. Instead of directly learning a meta-policy, we develop a Meta-learner with Subtask Graph Inference. Our experiment results on two grid-world domains and StarCraft II environments show that the proposed method is able to accurately infer the latent task parameter.
  arXiv  Detail & Related papers  (2020-01-01T17:34:00Z)

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.