Describe-and-Dissect: Interpreting Neurons in Vision Networks with Language Models

• URL: http://arxiv.org/abs/2403.13771v1
• Date: Wed, 20 Mar 2024 17:33:02 GMT
• Title: Describe-and-Dissect: Interpreting Neurons in Vision Networks with Language Models
• Authors: Nicholas Bai, Rahul A. Iyer, Tuomas Oikarinen, Tsui-Wei Weng,
• Abstract summary: Describe-and-Dissect (DnD) is a novel method to describe the roles of hidden neurons in vision networks. DnD produces complex natural language descriptions without the need for labeled training data or a predefined set of concepts.
• Score: 9.962488213825859
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: In this paper, we propose Describe-and-Dissect (DnD), a novel method to describe the roles of hidden neurons in vision networks. DnD utilizes recent advancements in multimodal deep learning to produce complex natural language descriptions, without the need for labeled training data or a predefined set of concepts to choose from. Additionally, DnD is training-free, meaning we don't train any new models and can easily leverage more capable general purpose models in the future. We have conducted extensive qualitative and quantitative analysis to show that DnD outperforms prior work by providing higher quality neuron descriptions. Specifically, our method on average provides the highest quality labels and is more than 2 times as likely to be selected as the best explanation for a neuron than the best baseline.

Related papers

• Towards Generating Informative Textual Description for Neurons in Language Models [6.884227665279812]
  We propose a framework that ties textual descriptions to neurons. In particular, our experiment shows that the proposed approach achieves 75% precision@2, and 50% recall@2
  arXiv  Detail & Related papers  (2024-01-30T04:06:25Z)
• Manipulating Feature Visualizations with Gradient Slingshots [54.31109240020007]
  We introduce a novel method for manipulating Feature Visualization (FV) without significantly impacting the model's decision-making process. We evaluate the effectiveness of our method on several neural network models and demonstrate its capabilities to hide the functionality of arbitrarily chosen neurons.
  arXiv  Detail & Related papers  (2024-01-11T18:57:17Z)
• Investigating the Encoding of Words in BERT's Neurons using Feature Textualization [11.943486282441143]
  We propose a technique to produce representations of neurons in embedding word space. We find that the produced representations can provide insights about the encoded knowledge in individual neurons.
  arXiv  Detail & Related papers  (2023-11-14T15:21:49Z)
• Automated Natural Language Explanation of Deep Visual Neurons with Large Models [43.178568768100305]
  This paper proposes a novel post-hoc framework for generating semantic explanations of neurons with large foundation models. Our framework is designed to be compatible with various model architectures and datasets, automated and scalable neuron interpretation.
  arXiv  Detail & Related papers  (2023-10-16T17:04:51Z)
• Evaluating Explanation Methods for Vision-and-Language Navigation [26.607847422870464]
  We propose a new erasure-based evaluation pipeline to measure the step-wise textual explanation in the sequential decision-making setting. We evaluate several explanation methods for two representative VLN models on two popular VLN datasets.
  arXiv  Detail & Related papers  (2023-10-10T14:22:56Z)
• Seeing in Words: Learning to Classify through Language Bottlenecks [59.97827889540685]
  Humans can explain their predictions using succinct and intuitive descriptions. We show that a vision model whose feature representations are text can effectively classify ImageNet images.
  arXiv  Detail & Related papers  (2023-06-29T00:24:42Z)
• Leveraging Graph-based Cross-modal Information Fusion for Neural Sign Language Translation [46.825957917649795]
  Sign Language (SL), as the mother tongue of the deaf community, is a special visual language that most hearing people cannot understand. We propose a novel neural SLT model with multi-modal feature fusion based on the dynamic graph. We are the first to introduce graph neural networks, for fusing multi-modal information, into neural sign language translation models.
  arXiv  Detail & Related papers  (2022-11-01T15:26:22Z)
• Natural Language Descriptions of Deep Visual Features [50.270035018478666]
  We introduce a procedure that automatically labels neurons with open-ended, compositional, natural language descriptions. We use MILAN for analysis, characterizing the distribution and importance of neurons selective for attribute, category, and relational information in vision models. We also use MILAN for auditing, surfacing neurons sensitive to protected categories like race and gender in models trained on datasets intended to obscure these features.
  arXiv  Detail & Related papers  (2022-01-26T18:48:02Z)
• Explainability Tools Enabling Deep Learning in Future In-Situ Real-Time Planetary Explorations [58.720142291102135]
  Deep learning (DL) has proven to be an effective machine learning and computer vision technique. Most of the Deep Neural Network (DNN) architectures are so complex that they are considered a 'black box' In this paper, we used integrated gradients to describe the attributions of each neuron to the output classes. It provides a set of explainability tools (ET) that opens the black box of a DNN so that the individual contribution of neurons to category classification can be ranked and visualized.
  arXiv  Detail & Related papers  (2022-01-15T07:10:00Z)
• Compositional Explanations of Neurons [52.71742655312625]
  We describe a procedure for explaining neurons in deep representations by identifying compositional logical concepts. We use this procedure to answer several questions on interpretability in models for vision and natural language processing.
  arXiv  Detail & Related papers  (2020-06-24T20:37:05Z)
• Non-linear Neurons with Human-like Apical Dendrite Activations [81.18416067005538]
  We show that a standard neuron followed by our novel apical dendrite activation (ADA) can learn the XOR logical function with 100% accuracy. We conduct experiments on six benchmark data sets from computer vision, signal processing and natural language processing.
  arXiv  Detail & Related papers  (2020-02-02T21:09:39Z)

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.