Loss-Aware Feature-Map Pruning in Convolutional Neural Networks Using Multi-Armed Bandits

arXiv cs.AI Papers

Summary

This paper introduces a loss-aware feature-map pruning framework for convolutional neural networks using multi-armed bandits (UCB1 and Thompson Sampling) to selectively remove redundant feature maps, reducing computational cost while preserving accuracy.

arXiv:2607.22564v1 Announce Type: new Abstract: Convolutional neural networks often contain redundant feature maps that increase storage and inference cost. This paper presents a loss-aware feature-map pruning framework using multi-armed bandits. Feature-map pruning is structured because it removes complete convolutional output channels and their producing filters rather than isolated scalar weights. Each candidate feature map is treated as an arm. At each play time, one map is temporarily masked and evaluated on a sampled mini-batch; the map is then restored and the observed loss change is converted into a safe-removal reward. After a fixed play budget, candidate maps are ranked by learned scores and the top-k maps are permanently removed with their filters, biases and corresponding next-layer input-channel kernels. The study evaluates UCB1 and Thompson Sampling, compares them with direct/oracle-style evaluation on LeNet/MNIST, and extends the evaluation to MNIST, CIFAR-10, CIFAR-100, SVHN, CUB-200-2011 and Oxford Flowers 102. Results show that UCB1 and Thompson Sampling preserve accuracy close to unpruned models while removing feature maps and reducing convolutional computation. Friedman and Nemenyi tests show that UCB1 obtains the highest mean rank, followed by Thompson Sampling; both significantly outperform greedy and magnitude-based pruning while remaining statistically comparable to the original unpruned model.
Original Article
View Cached Full Text

Cached at: 07/28/26, 06:24 AM

# Loss-Aware Feature-Map Pruning in Convolutional Neural Networks Using Multi-Armed Bandits
Source: [https://arxiv.org/abs/2607.22564](https://arxiv.org/abs/2607.22564)
[View PDF](https://arxiv.org/pdf/2607.22564)

> Abstract:Convolutional neural networks often contain redundant feature maps that increase storage and inference cost\. This paper presents a loss\-aware feature\-map pruning framework using multi\-armed bandits\. Feature\-map pruning is structured because it removes complete convolutional output channels and their producing filters rather than isolated scalar weights\. Each candidate feature map is treated as an arm\. At each play time, one map is temporarily masked and evaluated on a sampled mini\-batch; the map is then restored and the observed loss change is converted into a safe\-removal reward\. After a fixed play budget, candidate maps are ranked by learned scores and the top\-k maps are permanently removed with their filters, biases and corresponding next\-layer input\-channel kernels\. The study evaluates UCB1 and Thompson Sampling, compares them with direct/oracle\-style evaluation on LeNet/MNIST, and extends the evaluation to MNIST, CIFAR\-10, CIFAR\-100, SVHN, CUB\-200\-2011 and Oxford Flowers 102\. Results show that UCB1 and Thompson Sampling preserve accuracy close to unpruned models while removing feature maps and reducing convolutional computation\. Friedman and Nemenyi tests show that UCB1 obtains the highest mean rank, followed by Thompson Sampling; both significantly outperform greedy and magnitude\-based pruning while remaining statistically comparable to the original unpruned model\.

## Submission history

From: Salem Ameen \[[view email](https://arxiv.org/show-email/4f89522d/2607.22564)\] **\[v1\]**Fri, 29 May 2026 20:03:08 UTC \(560 KB\)

Similar Articles

An AI4AI Framework for Visual Token Pruning

Hugging Face Daily Papers

AutoPrune is a training-free framework that uses LLMs to automatically design visual-token pruning policies for multimodal LLMs via a domain-specific language and residual search, achieving high efficiency with minimal performance loss (99% performance retained while removing 94.4% of visual tokens).