Efficient Data Driven Mixture-of-Expert Extraction from Trained Networks

• URL: http://arxiv.org/abs/2505.15414v1
• Date: Wed, 21 May 2025 11:55:25 GMT
• Title: Efficient Data Driven Mixture-of-Expert Extraction from Trained Networks
• Abstract summary: Vision Transformers have emerged as the state-of-the-art models in various Computer Vision tasks.<n>They often require costly retraining or even training from scratch.<n>Recent developments aim to reduce these computational costs by leveraging pretrained networks.
• Score: 46.498278084317704
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Vision Transformers have emerged as the state-of-the-art models in various Computer Vision tasks, but their high computational and resource demands pose significant challenges. While Mixture-of-Experts (MoE) can make these models more efficient, they often require costly retraining or even training from scratch. Recent developments aim to reduce these computational costs by leveraging pretrained networks. These have been shown to produce sparse activation patterns in the Multi-Layer Perceptrons (MLPs) of the encoder blocks, allowing for conditional activation of only relevant subnetworks for each sample. Building on this idea, we propose a new method to construct MoE variants from pretrained models. Our approach extracts expert subnetworks from the model's MLP layers post-training in two phases. First, we cluster output activations to identify distinct activation patterns. In the second phase, we use these clusters to extract the corresponding subnetworks responsible for producing them. On ImageNet-1k recognition tasks, we demonstrate that these extracted experts can perform surprisingly well out of the box and require only minimal fine-tuning to regain 98% of the original performance, all while reducing MACs and model size, by up to 36% and 32% respectively.

Related papers

• Explorative Modeling: Unlocking a Third Pretraining Axis and End-to-End Generation [89.51408170909548]
  We introduce Explorative Modeling, a new paradigm that factors the training loop, exploring K candidate matches between model generations and data, and training on the best.<n>First, increasing exploration adds a third pretraining axis beyond parameters and data for existing generative models.<n>Second, XMs enable end-to-end reconstructive generative modeling, matching diffusion on control tasks with 16-256x fewer inference steps.
  arXiv  Detail & Related papers  (2026-07-29T18:25:17Z)
• Recursive Block-Diagonal Coupling for Resource-Efficient Training of Vision Models [16.54656665398701]
  Training high-capacity vision models from scratch requires substantial computational resources.<n>We propose an efficient training protocol, RBDC, that builds wide models by coupling in a parameter-free block-diagonal way.<n>Our protocol shows a much better efficiency than models trained from scratch with the standard protocol, yielding 30% FLOPs reduction at similar test accuracies.
  arXiv  Detail & Related papers  (2026-05-22T14:08:22Z)
• Effective Data Pruning through Score Extrapolation [40.61665742457229]
  We introduce a novel importance score extrapolation framework that requires training on only a small subset of data.<n>We present two initial approaches in this framework to accurately predict sample importance for the entire dataset using patterns learned from this minimal subset.<n>Our results indicate that score extrapolation is a promising direction to scale expensive score calculation methods, such as pruning, data attribution, or other tasks.
  arXiv  Detail & Related papers  (2025-06-10T17:38:49Z)
• Is Tokenization Needed for Masked Particle Modelling? [8.79008927474707]
  Masked particle modeling (MPM) is a self-supervised learning scheme for constructing expressive representations of unordered sets. We improve MPM by addressing inefficiencies in the implementation and incorporating a more powerful decoder. We show that these new methods outperform the tokenized learning objective from the original MPM on a new test bed for foundation models for jets.
  arXiv  Detail & Related papers  (2024-09-19T09:12:29Z)
• Asymmetric Masked Distillation for Pre-Training Small Foundation Models [52.56257450614992]
  Self-supervised foundation models have shown great potential in computer vision thanks to the pre-training paradigm of masked autoencoding. This paper focuses on pre-training relatively small vision transformer models that could be efficiently adapted to downstream tasks. We propose a new asymmetric masked distillation (AMD) framework for pre-training relatively small models with autoencoding.
  arXiv  Detail & Related papers  (2023-11-06T14:44:34Z)
• LEMON: Lossless model expansion [43.40389747029802]
  Scaling of deep neural networks, especially Transformers, is pivotal for their surging performance. We present $textbfL$ossl$textbfE$ss $textbfMO$del Expansio$textbfN$ (LEMON), a recipe to initialize scaled models using the weights of their smaller but pre-trained counterparts. We show that LEMON reduces computational costs by 56.7% for Vision Transformers and 33.2% for BERT when compared to training from scratch.
  arXiv  Detail & Related papers  (2023-10-12T03:02:41Z)
• Sheared LLaMA: Accelerating Language Model Pre-training via Structured Pruning [52.29522018586365]
  We study structured pruning as an effective means to develop smaller LLMs from pre-trained, larger models. Our approach employs two key techniques: (1) targeted structured pruning, which prunes a larger model to a specified target shape by removing layers, heads, and intermediate and hidden dimensions in an end-to-end manner, and (2) dynamic batch loading, which dynamically updates the composition of sampled data in each training batch based on varying losses across different domains.
  arXiv  Detail & Related papers  (2023-10-10T15:13:30Z)
• 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)
• Learning Energy-Based Models by Cooperative Diffusion Recovery Likelihood [64.95663299945171]
  Training energy-based models (EBMs) on high-dimensional data can be both challenging and time-consuming. There exists a noticeable gap in sample quality between EBMs and other generative frameworks like GANs and diffusion models. We propose cooperative diffusion recovery likelihood (CDRL), an effective approach to tractably learn and sample from a series of EBMs.
  arXiv  Detail & Related papers  (2023-09-10T22:05:24Z)
• One-Shot Pruning for Fast-adapting Pre-trained Models on Devices [28.696989086706186]
  Large-scale pre-trained models have been remarkably successful in resolving downstream tasks. deploying these models on low-capability devices still requires an effective approach, such as model pruning. We present a scalable one-shot pruning method that leverages pruned knowledge of similar tasks to extract a sub-network from the pre-trained model for a new task.
  arXiv  Detail & Related papers  (2023-07-10T06:44:47Z)
• Efficient Deep Spiking Multi-Layer Perceptrons with Multiplication-Free Inference [13.924924047051782]
  Deep convolution architectures for Spiking Neural Networks (SNNs) have significantly enhanced image classification performance and reduced computational burdens. This research explores a new pathway, drawing inspiration from the progress made in Multi-Layer Perceptrons (MLPs) We propose an innovative spiking architecture that uses batch normalization to retain MFI compatibility. We establish an efficient multi-stage spiking network that blends effectively global receptive fields with local feature extraction.
  arXiv  Detail & Related papers  (2023-06-21T16:52:20Z)
• Voting from Nearest Tasks: Meta-Vote Pruning of Pre-trained Models for Downstream Tasks [55.431048995662714]
  We create a small model for a new task from the pruned models of similar tasks. We show that a few fine-tuning steps on this model suffice to produce a promising pruned-model for the new task. We develop a simple but effective ''Meta-Vote Pruning (MVP)'' method that significantly reduces the pruning iterations for a new task.
  arXiv  Detail & Related papers  (2023-01-27T06:49:47Z)
• MoEBERT: from BERT to Mixture-of-Experts via Importance-Guided Adaptation [68.30497162547768]
  We propose MoEBERT, which uses a Mixture-of-Experts structure to increase model capacity and inference speed. We validate the efficiency and effectiveness of MoEBERT on natural language understanding and question answering tasks.
  arXiv  Detail & Related papers  (2022-04-15T23:19:37Z)

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.