CP-MoE: Consistency-Preserving Mixture-of-Experts for Continual Learning

• URL: http://arxiv.org/abs/2605.20247v1
• Date: Mon, 18 May 2026 06:31:14 GMT
• Title: CP-MoE: Consistency-Preserving Mixture-of-Experts for Continual Learning
• Abstract summary: Catastrophic forgetting remains a major obstacle to continual learning in large language models (LLMs) and vision--language models (VLMs)<n>We propose CP-MoE, a continual learning framework built around a transient expert that captures early task-specific updates and guides their integration into stable experts.<n>We validate CP-MoE on both unimodal and multimodal continual learning benchmarks with LLM-based and VLM-based MoE models.
• Score: 16.033361968793162
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Catastrophic forgetting remains a major obstacle to continual learning in large language models (LLMs) and vision--language models (VLMs). Although Mixture-of-Experts (MoE) architectures offer an efficient path to scaling, existing LoRA-based MoE continual learning methods still face a fundamental trade-off: they either isolate experts too aggressively, limiting knowledge transfer across tasks, or allow task-specific updates to overwrite important existing parameters, leading to severe forgetting. To address this, we propose CP-MoE, a continual learning framework built around a transient expert that captures early task-specific updates and guides their integration into stable experts. CP-MoE introduces a consistency-preserving routing bias, which uses the transient expert to estimate representation similarity with stable experts and steer routing towards more compatible expert selection, and a transient expert-guided regularisation mechanism, which selectively protects important historical parameters during merging. Together, these components reduce parameter interference and forgetting while preserving cross-task knowledge transfer. We validate CP-MoE on both unimodal and multimodal continual learning benchmarks with LLM-based and VLM-based MoE models. On SuperNI benchmark, spanning diverse sequential language tasks, CP-MoE achieves state-of-the-art performance and stronger zero-shot transfer to unseen tasks. On VQA v2 dataset, it scales effectively to multimodal visual reasoning, consistently reduces forgetting, and outperforms strong MoE baselines.

Related papers

• Tying the Loop -- Tied Expert Layers in Mixture-of-Experts Language Models [50.34585122347149]
  We introduce Expert Tying, an architectural modification that shares expert parameters across consecutive transformer layers.<n>We evaluate this approach across common, state-of-the-art architectures, including OLMoE, Qwen3, and DeepSeek-style MoEs.
  arXiv  Detail & Related papers  (2026-06-15T15:08:09Z)
• FutureMind: Equipping Small Language Models with Strategic Thinking-Pattern Priors via Adaptive Knowledge Distillation [13.855534865501369]
  Small Language Models (SLMs) are attractive for cost-sensitive and resource-limited settings due to their efficient, low-latency inference.<n>We propose FutureMind, a modular reasoning framework that equips SLMs with strategic thinking-pattern priors.
  arXiv  Detail & Related papers  (2026-02-01T13:26:04Z)
• From Sparse Decisions to Dense Reasoning: A Multi-attribute Trajectory Paradigm for Multimodal Moderation [59.27094165576015]
  We propose a novel learning paradigm (UniMod) that transitions from sparse decision-making to dense reasoning traces.<n>By constructing structured trajectories encompassing evidence grounding, modality assessment, risk mapping, policy decision, and response generation, we reformulate monolithic decision tasks into a multi-dimensional boundary learning process.<n>We introduce specialized optimization strategies to decouple task-specific parameters and rebalance training dynamics, effectively resolving interference between diverse objectives in multi-task learning.
  arXiv  Detail & Related papers  (2026-01-28T09:29:40Z)
• AR-MOT: Autoregressive Multi-object Tracking [56.09738000988466]
  We propose a novel autoregressive paradigm that formulates MOT as a sequence generation task within a large language model (LLM) framework.<n>This design enables the model to output structured results through flexible sequence construction, without requiring any task-specific heads.<n>To enhance region-level visual perception, we introduce an Object Tokenizer based on a pretrained detector.
  arXiv  Detail & Related papers  (2026-01-05T09:17:28Z)
• Mixtures of SubExperts for Large Language Continual Learning [6.425296129700846]
  Adapting Large Language Models to a continuous stream of tasks is a critical yet challenging endeavor.<n>Reusing a single set of PEFT parameters for new tasks often leads to catastrophic forgetting of prior knowledge.<n>We propose a novel adaptive PEFT method referred to as textitMixtures of SubExperts (MoSEs), a novel continual learning framework designed for minimal forgetting and efficient scalability.
  arXiv  Detail & Related papers  (2025-11-09T05:44:45Z)
• Continual Learning for VLMs: A Survey and Taxonomy Beyond Forgetting [70.83781268763215]
  Vision-language models (VLMs) have achieved impressive performance across diverse multimodal tasks by leveraging large-scale pre-training.<n>VLMs face unique challenges such as cross-modal feature drift, parameter interference due to shared architectures, and zero-shot capability erosion.<n>This survey aims to serve as a comprehensive and diagnostic reference for researchers developing lifelong vision-language systems.
  arXiv  Detail & Related papers  (2025-08-06T09:03:10Z)
• Bisecle: Binding and Separation in Continual Learning for Video Language Understanding [11.710573955384511]
  We propose Bisecle for video-language continual learning, inspired by the rapid Binding and pattern separation mechanisms in the hippocampus.<n>Bisecle mitigates forgetting and enhances cross-task generalization on several VideoQA benchmarks.
  arXiv  Detail & Related papers  (2025-07-01T06:28:57Z)
• MoCa: Modality-aware Continual Pre-training Makes Better Bidirectional Multimodal Embeddings [75.0617088717528]
  MoCa is a framework for transforming pre-trained VLM backbones into effective bidirectional embedding models.<n>MoCa consistently improves performance across MMEB and ViDoRe-v2 benchmarks, achieving new state-of-the-art results.
  arXiv  Detail & Related papers  (2025-06-29T06:41:00Z)
• LLaVA-CMoE: Towards Continual Mixture of Experts for Large Vision-Language Models [21.888139819188105]
  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.
  arXiv  Detail & Related papers  (2025-03-27T07:36:11Z)
• CL-MoE: Enhancing Multimodal Large Language Model with Dual Momentum Mixture-of-Experts for Continual Visual Question Answering [27.812611421754482]
  We propose an MLLMs-based dual momentum Mixture-of-Experts (CL-MoE) framework for continual visual question answering (VQA)<n>We integrate MLLMs with continual learning to utilize the rich commonsense knowledge in LLMs.<n>Our method achieves state-of-the-art performance on 10 VQA tasks, proving the effectiveness of our approach.
  arXiv  Detail & Related papers  (2025-03-01T09:25:23Z)
• FactorLLM: Factorizing Knowledge via Mixture of Experts for Large Language Models [50.331708897857574]
  We introduce FactorLLM, a novel approach that decomposes well-trained dense FFNs into sparse sub-networks without requiring any further modifications. FactorLLM achieves comparable performance to the source model securing up to 85% model performance while obtaining over a 30% increase in inference speed.
  arXiv  Detail & Related papers  (2024-08-15T16:45:16Z)
• Boosting Continual Learning of Vision-Language Models via Mixture-of-Experts Adapters [65.15700861265432]
  We present a parameter-efficient continual learning framework to alleviate long-term forgetting in incremental learning with vision-language models. Our approach involves the dynamic expansion of a pre-trained CLIP model, through the integration of Mixture-of-Experts (MoE) adapters. To preserve the zero-shot recognition capability of vision-language models, we introduce a Distribution Discriminative Auto-Selector.
  arXiv  Detail & Related papers  (2024-03-18T08:00:23Z)

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.