Gradient Routing: Masking Gradients to Localize Computation in Neural Networks

• URL: http://arxiv.org/abs/2410.04332v1
• Date: Sun, 6 Oct 2024 02:43:49 GMT
• Title: Gradient Routing: Masking Gradients to Localize Computation in Neural Networks
• Authors: Alex Cloud, Jacob Goldman-Wetzler, Evžen Wybitul, Joseph Miller, Alexander Matt Turner,
• Abstract summary: We introduce gradient routing, a training method that isolates capabilities to specific subregions of a neural network. We show that gradient routing can be used to learn representations which are partitioned in an interpretable way. We conclude that the approach holds promise for challenging, real-world applications where quality data are scarce.
• Score: 43.0686937643683
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Neural networks are trained primarily based on their inputs and outputs, without regard for their internal mechanisms. These neglected mechanisms determine properties that are critical for safety, like (i) transparency; (ii) the absence of sensitive information or harmful capabilities; and (iii) reliable generalization of goals beyond the training distribution. To address this shortcoming, we introduce gradient routing, a training method that isolates capabilities to specific subregions of a neural network. Gradient routing applies data-dependent, weighted masks to gradients during backpropagation. These masks are supplied by the user in order to configure which parameters are updated by which data points. We show that gradient routing can be used to (1) learn representations which are partitioned in an interpretable way; (2) enable robust unlearning via ablation of a pre-specified network subregion; and (3) achieve scalable oversight of a reinforcement learner by localizing modules responsible for different behaviors. Throughout, we find that gradient routing localizes capabilities even when applied to a limited, ad-hoc subset of the data. We conclude that the approach holds promise for challenging, real-world applications where quality data are scarce.

Related papers

• Neural Network Pruning by Gradient Descent [7.427858344638741]
  We introduce a novel and straightforward neural network pruning framework that incorporates the Gumbel-Softmax technique. We demonstrate its exceptional compression capability, maintaining high accuracy on the MNIST dataset with only 0.15% of the original network parameters. We believe our method opens a promising new avenue for deep learning pruning and the creation of interpretable machine learning systems.
  arXiv  Detail & Related papers  (2023-11-21T11:12:03Z)
• A General Framework for Interpretable Neural Learning based on Local Information-Theoretic Goal Functions [1.5236380958983644]
  We introduce 'infomorphic' neural networks to perform tasks from supervised, unsupervised and memory learning. By leveraging the interpretable nature of the PID framework, infomorphic networks represent a valuable tool to advance our understanding of the intricate structure of local learning.
  arXiv  Detail & Related papers  (2023-06-03T16:34:25Z)
• Inducing Gaussian Process Networks [80.40892394020797]
  We propose inducing Gaussian process networks (IGN), a simple framework for simultaneously learning the feature space as well as the inducing points. The inducing points, in particular, are learned directly in the feature space, enabling a seamless representation of complex structured domains. We report on experimental results for real-world data sets showing that IGNs provide significant advances over state-of-the-art methods.
  arXiv  Detail & Related papers  (2022-04-21T05:27:09Z)
• Improving Deep Learning Interpretability by Saliency Guided Training [36.782919916001624]
  Saliency methods have been widely used to highlight important input features in model predictions. Most existing methods use backpropagation on a modified gradient function to generate saliency maps. We introduce a saliency guided training procedure for neural networks to reduce noisy gradients used in predictions.
  arXiv  Detail & Related papers  (2021-11-29T06:05:23Z)
• Coarse to Fine: Domain Adaptive Crowd Counting via Adversarial Scoring Network [58.05473757538834]
  This paper proposes a novel adversarial scoring network (ASNet) to bridge the gap across domains from coarse to fine granularity. Three sets of migration experiments show that the proposed methods achieve state-of-the-art counting performance.
  arXiv  Detail & Related papers  (2021-07-27T14:47:24Z)
• Unsupervised Metric Relocalization Using Transform Consistency Loss [66.19479868638925]
  Training networks to perform metric relocalization traditionally requires accurate image correspondences. We propose a self-supervised solution, which exploits a key insight: localizing a query image within a map should yield the same absolute pose, regardless of the reference image used for registration. We evaluate our framework on synthetic and real-world data, showing our approach outperforms other supervised methods when a limited amount of ground-truth information is available.
  arXiv  Detail & Related papers  (2020-11-01T19:24:27Z)
• Layer-wise Characterization of Latent Information Leakage in Federated Learning [9.397152006395174]
  Training deep neural networks via federated learning allows clients to share, instead of the original data, only the model trained on their data. Prior work has demonstrated that in practice a client's private information, unrelated to the main learning task, can be discovered from the model's gradients. There is still no formal approach for quantifying the leakage of private information via the shared updated model or gradients.
  arXiv  Detail & Related papers  (2020-10-17T10:49:14Z)
• Counterfactual Explanation Based on Gradual Construction for Deep Networks [17.79934085808291]
  The patterns that deep networks have learned from a training dataset can be grasped by observing the feature variation among various classes. Current approaches perform the feature modification to increase the classification probability for the target class irrespective of the internal characteristics of deep networks. We propose a counterfactual explanation method that exploits the statistics learned from a training dataset.
  arXiv  Detail & Related papers  (2020-08-05T01:18:31Z)
• Neural networks adapting to datasets: learning network size and topology [77.34726150561087]
  We introduce a flexible setup allowing for a neural network to learn both its size and topology during the course of a gradient-based training. The resulting network has the structure of a graph tailored to the particular learning task and dataset.
  arXiv  Detail & Related papers  (2020-06-22T12:46:44Z)
• Understanding Self-Training for Gradual Domain Adaptation [107.37869221297687]
  We consider gradual domain adaptation, where the goal is to adapt an initial classifier trained on a source domain given only unlabeled data that shifts gradually in distribution towards a target domain. We prove the first non-vacuous upper bound on the error of self-training with gradual shifts, under settings where directly adapting to the target domain can result in unbounded error. The theoretical analysis leads to algorithmic insights, highlighting that regularization and label sharpening are essential even when we have infinite data, and suggesting that self-training works particularly well for shifts with small Wasserstein-infinity distance.
  arXiv  Detail & Related papers  (2020-02-26T08:59:40Z)

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.