# Analyzing and Mitigating Interference in Neural Architecture Search

arXiv: https://arxiv.org/abs/2108.12821 · HTML: https://mlsanity.com/paper/2108.12821/

## Key facts

- Cited by 20 papers in the map (data through 7 Oct 2026).
- Builds on 61 of its 77 references that are in the map.
- First posted on arXiv 29 Aug 2021.

## Builds on (most cited first, 30 of 61)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://mlsanity.com/paper/1907.11692/) (2019)
- [mixup: Beyond Empirical Risk Minimization](https://mlsanity.com/paper/1710.09412/) (2017)
- [GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language Understanding](https://mlsanity.com/paper/1804.07461/) (2018)
- [MobileNetV2: Inverted Residuals and Linear Bottlenecks](https://mlsanity.com/paper/1801.04381/) (2018)
- [Squeeze-and-Excitation Networks](https://mlsanity.com/paper/1709.01507/) (2017)
- [DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter](https://mlsanity.com/paper/1910.01108/) (2019)
- [SQuAD: 100,000+ Questions for Machine Comprehension of Text](https://mlsanity.com/paper/1606.05250/) (2016)
- [ALBERT: A Lite BERT for Self-supervised Learning of Language Representations](https://mlsanity.com/paper/1909.11942/) (2019)
- [A Broad-Coverage Challenge Corpus for Sentence Understanding through Inference](https://mlsanity.com/paper/1704.05426/) (2017)
- [Xception: Deep Learning with Depthwise Separable Convolutions](https://mlsanity.com/paper/1610.02357/) (2016)
- [DARTS: Differentiable Architecture Search](https://mlsanity.com/paper/1806.09055/) (2018)
- [Improved Regularization of Convolutional Neural Networks with Cutout](https://mlsanity.com/paper/1708.04552/) (2017)
- [Know What You Don't Know: Unanswerable Questions for SQuAD](https://mlsanity.com/paper/1806.03822/) (2018)
- [Neural Architecture Search with Reinforcement Learning](https://mlsanity.com/paper/1611.01578/) (2016)
- [fairseq: A Fast, Extensible Toolkit for Sequence Modeling](https://mlsanity.com/paper/1904.01038/) (2019)
- [MiniLM: Deep Self-Attention Distillation for Task-Agnostic Compression of Pre-Trained Transformers](https://mlsanity.com/paper/2002.10957/) (2020)
- [ShuffleNet: An Extremely Efficient Convolutional Neural Network for Mobile Devices](https://mlsanity.com/paper/1707.01083/) (2017)
- [TinyBERT: Distilling BERT for Natural Language Understanding](https://mlsanity.com/paper/1909.10351/) (2019)
- [Learning Transferable Architectures for Scalable Image Recognition](https://mlsanity.com/paper/1707.07012/) (2017)
- [SemEval-2017 Task 1: Semantic Textual Similarity - Multilingual and Cross-lingual Focused Evaluation](https://mlsanity.com/paper/1708.00055/) (2017)
- [MPNet: Masked and Permuted Pre-training for Language Understanding](https://mlsanity.com/paper/2004.09297/) (2020)
- [Neural Network Acceptability Judgments](https://mlsanity.com/paper/1805.12471/) (2018)
- [Efficient Neural Architecture Search via Parameter Sharing](https://mlsanity.com/paper/1802.03268/) (2018)
- [Once-for-All: Train One Network and Specialize it for Efficient Deployment](https://mlsanity.com/paper/1908.09791/) (2019)
- [Unified Language Model Pre-training for Natural Language Understanding and Generation](https://mlsanity.com/paper/1905.03197/) (2019)
- [ProxylessNAS: Direct Neural Architecture Search on Target Task and Hardware](https://mlsanity.com/paper/1812.00332/) (2018)
- [Patient Knowledge Distillation for BERT Model Compression](https://mlsanity.com/paper/1908.09355/) (2019)
- [Single Path One-Shot Neural Architecture Search with Uniform Sampling](https://mlsanity.com/paper/1904.00420/) (2019)

## Built on by (most cited first, 20 of 20)

- [Neural Architecture Search: Insights from 1000 Papers](https://mlsanity.com/paper/2301.08727/) (2023)
- [ZiCo: Zero-shot NAS via Inverse Coefficient of Variation on Gradients](https://mlsanity.com/paper/2301.11300/) (2023)
- [LLM Performance Predictors are good initializers for Architecture Search](https://mlsanity.com/paper/2310.16712/) (2023)
- [LiteTransformerSearch: Training-free Neural Architecture Search for Efficient Language Models](https://mlsanity.com/paper/2203.02094/) (2022)
- [ElasticViT: Conflict-aware Supernet Training for Deploying Fast Vision Transformer on Diverse Mobile Devices](https://mlsanity.com/paper/2303.09730/) (2023)
- [AutoMoE: Heterogeneous Mixture-of-Experts with Adaptive Computation for Efficient Neural Machine Translation](https://mlsanity.com/paper/2210.07535/) (2022)
- [Mixture-of-Supernets: Improving Weight-Sharing Supernet Training with Architecture-Routed Mixture-of-Experts](https://mlsanity.com/paper/2306.04845/) (2023)
- [PA&DA: Jointly Sampling PAth and DAta for Consistent NAS](https://mlsanity.com/paper/2302.14772/) (2023)
- [NAS-LID: Efficient Neural Architecture Search with Local Intrinsic Dimension](https://mlsanity.com/paper/2211.12759/) (2022)
- [FemtoDet: An Object Detection Baseline for Energy Versus Performance Tradeoffs](https://mlsanity.com/paper/2301.06719/) (2023)
- [Weight-Entanglement Meets Gradient-Based Neural Architecture Search](https://mlsanity.com/paper/2312.10440/) (2023)
- [Boosting Order-Preserving and Transferability for Neural Architecture Search: a Joint Architecture Refined Search and Fine-tuning Approach](https://mlsanity.com/paper/2403.11380/) (2024)
- [Automated Dominative Subspace Mining for Efficient Neural Architecture Search](https://mlsanity.com/paper/2210.17180/) (2022)
- [LayerNAS: Neural Architecture Search in Polynomial Complexity](https://mlsanity.com/paper/2304.11517/) (2023)
- [Evolutionary Neural Architecture Search for Transformer in Knowledge Tracing](https://mlsanity.com/paper/2310.01180/) (2023)
- [A Survey of Supernet Optimization and its Applications: Spatial and Temporal Optimization for Neural Architecture Search](https://mlsanity.com/paper/2204.03916/) (2022)
- [AutoHEnsGNN: Winning Solution to AutoGraph Challenge for KDD Cup 2020](https://mlsanity.com/paper/2111.12952/) (2021)
- [Does Interference Exist When Training a Once-For-All Network?](https://mlsanity.com/paper/2204.09210/) (2022)
- [NAS-FM: Neural Architecture Search for Tunable and Interpretable Sound Synthesis based on Frequency Modulation](https://mlsanity.com/paper/2305.12868/) (2023)
- [InTrain: Intrinsic Trainability for Zero-Cost Neural Architecture Search](https://mlsanity.com/paper/2606.18676/) (2026)

## Abstract (arXiv)

Weight sharing is a popular approach to reduce the cost of neural architecture search (NAS) by reusing the weights of shared operators from previously trained child models. However, the rank correlation between the estimated accuracy and ground truth accuracy of those child models is low due to the interference among different child models caused by weight sharing. In this paper, we investigate the interference issue by sampling different child models and calculating the gradient similarity of shared operators, and observe: 1) the interference on a shared operator between two child models is positively correlated with the number of different operators; 2) the interference is smaller when the inputs and outputs of the shared operator are more similar. Inspired by these two observations, we propose two approaches to mitigate the interference: 1) MAGIC-T: rather than randomly sampling child models for optimization, we propose a gradual modification scheme by modifying one operator between adjacent optimization steps to minimize the interference on the shared operators; 2) MAGIC-A: forcing the inputs and outputs of the operator across all child models to be similar to reduce the interference. Experiments on a BERT search space verify that mitigating interference via each of our proposed methods improves the rank correlation of super-pet and combining both methods can achieve better results. Our discovered architecture outperforms RoBERTarm base by 1.1 and 0.6 points and ELECTRArm base by 1.6 and 1.1 points on the dev and test set of GLUE benchmark. Extensive results on the BERT compression, reading comprehension and ImageNet task demonstrate the effectiveness and generality of our proposed methods.