Concept-Based Mechanistic Interpretability Using Structured Knowledge Graphs

• URL: http://arxiv.org/abs/2507.05810v1
• Date: Tue, 08 Jul 2025 09:30:20 GMT
• Title: Concept-Based Mechanistic Interpretability Using Structured Knowledge Graphs
• Authors: Sofiia Chorna, Kateryna Tarelkina, Eloïse Berthier, Gianni Franchi,
• Abstract summary: Our framework enables a global dissection of model behavior by analyzing how high-level semantic attributes emerge, interact, and propagate through internal model components.<n>A key innovation is our visualization platform that we named BAGEL, which presents these insights in a structured knowledge graph.<n>Our framework is model-agnostic, scalable, and contributes to a deeper understanding of how deep learning models generalize (or fail to) in the presence of dataset biases.
• Score: 3.429783703166407
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: While concept-based interpretability methods have traditionally focused on local explanations of neural network predictions, we propose a novel framework and interactive tool that extends these methods into the domain of mechanistic interpretability. Our approach enables a global dissection of model behavior by analyzing how high-level semantic attributes (referred to as concepts) emerge, interact, and propagate through internal model components. Unlike prior work that isolates individual neurons or predictions, our framework systematically quantifies how semantic concepts are represented across layers, revealing latent circuits and information flow that underlie model decision-making. A key innovation is our visualization platform that we named BAGEL (for Bias Analysis with a Graph for global Explanation Layers), which presents these insights in a structured knowledge graph, allowing users to explore concept-class relationships, identify spurious correlations, and enhance model trustworthiness. Our framework is model-agnostic, scalable, and contributes to a deeper understanding of how deep learning models generalize (or fail to) in the presence of dataset biases. The demonstration is available at https://knowledge-graph-ui-4a7cb5.gitlab.io/.

Related papers

• Concept-Guided Interpretability via Neural Chunking [54.73787666584143]
  We show that neural networks exhibit patterns in their raw population activity that mirror regularities in the training data.<n>We propose three methods to extract these emerging entities, complementing each other based on label availability and dimensionality.<n>Our work points to a new direction for interpretability, one that harnesses both cognitive principles and the structure of naturalistic data.
  arXiv  Detail & Related papers  (2025-05-16T13:49:43Z)
• A Comprehensive Survey on Self-Interpretable Neural Networks [36.0575431131253]
  Self-interpretable neural networks inherently reveal the prediction rationale through the model structures.<n>We first collect and review existing works on self-interpretable neural networks and provide a structured summary of their methodologies.<n>We also present concrete, visualized examples of model explanations and discuss their applicability across diverse scenarios.
  arXiv  Detail & Related papers  (2025-01-26T18:50:16Z)
• Deep Learning Through A Telescoping Lens: A Simple Model Provides Empirical Insights On Grokking, Gradient Boosting & Beyond [61.18736646013446]
  In pursuit of a deeper understanding of its surprising behaviors, we investigate the utility of a simple yet accurate model of a trained neural network. Across three case studies, we illustrate how it can be applied to derive new empirical insights on a diverse range of prominent phenomena.
  arXiv  Detail & Related papers  (2024-10-31T22:54:34Z)
• Interpreting Inflammation Prediction Model via Tag-based Cohort Explanation [5.356481722174994]
  We propose a novel framework for identifying cohorts within a dataset based on local feature importance scores. We evaluate our framework on a food-based inflammation prediction model and demonstrated that the framework can generate reliable explanations that match domain knowledge.
  arXiv  Detail & Related papers  (2024-10-17T23:22:59Z)
• Decompose the model: Mechanistic interpretability in image models with Generalized Integrated Gradients (GIG) [24.02036048242832]
  This paper introduces a novel approach to trace the entire pathway from input through all intermediate layers to the final output within the whole dataset. We utilize Pointwise Feature Vectors (PFVs) and Effective Receptive Fields (ERFs) to decompose model embeddings into interpretable Concept Vectors. Then, we calculate the relevance between concept vectors with our Generalized Integrated Gradients (GIG) enabling a comprehensive, dataset-wide analysis of model behavior.
  arXiv  Detail & Related papers  (2024-09-03T05:19:35Z)
• Advancing Ante-Hoc Explainable Models through Generative Adversarial Networks [24.45212348373868]
  This paper presents a novel concept learning framework for enhancing model interpretability and performance in visual classification tasks. Our approach appends an unsupervised explanation generator to the primary classifier network and makes use of adversarial training. This work presents a significant step towards building inherently interpretable deep vision models with task-aligned concept representations.
  arXiv  Detail & Related papers  (2024-01-09T16:16:16Z)
• On the Joint Interaction of Models, Data, and Features [82.60073661644435]
  We introduce a new tool, the interaction tensor, for empirically analyzing the interaction between data and model through features. Based on these observations, we propose a conceptual framework for feature learning. Under this framework, the expected accuracy for a single hypothesis and agreement for a pair of hypotheses can both be derived in closed-form.
  arXiv  Detail & Related papers  (2023-06-07T21:35:26Z)
• Concept-Centric Transformers: Enhancing Model Interpretability through Object-Centric Concept Learning within a Shared Global Workspace [1.6574413179773757]
  Concept-Centric Transformers is a simple yet effective configuration of the shared global workspace for interpretability. We show that our model achieves better classification accuracy than all baselines across all problems.
  arXiv  Detail & Related papers  (2023-05-25T06:37:39Z)
• NxPlain: Web-based Tool for Discovery of Latent Concepts [16.446370662629555]
  We present NxPlain, a web application that provides an explanation of a model's prediction using latent concepts. NxPlain discovers latent concepts learned in a deep NLP model, provides an interpretation of the knowledge learned in the model, and explains its predictions based on the used concepts.
  arXiv  Detail & Related papers  (2023-03-06T10:45:24Z)
• Dynamic Inference with Neural Interpreters [72.90231306252007]
  We present Neural Interpreters, an architecture that factorizes inference in a self-attention network as a system of modules. inputs to the model are routed through a sequence of functions in a way that is end-to-end learned. We show that Neural Interpreters perform on par with the vision transformer using fewer parameters, while being transferrable to a new task in a sample efficient manner.
  arXiv  Detail & Related papers  (2021-10-12T23:22:45Z)
• Edge-assisted Democratized Learning Towards Federated Analytics [67.44078999945722]
  We show the hierarchical learning structure of the proposed edge-assisted democratized learning mechanism, namely Edge-DemLearn. We also validate Edge-DemLearn as a flexible model training mechanism to build a distributed control and aggregation methodology in regions.
  arXiv  Detail & Related papers  (2020-12-01T11:46:03Z)
• Concept Learners for Few-Shot Learning [76.08585517480807]
  We propose COMET, a meta-learning method that improves generalization ability by learning to learn along human-interpretable concept dimensions. We evaluate our model on few-shot tasks from diverse domains, including fine-grained image classification, document categorization and cell type annotation.
  arXiv  Detail & Related papers  (2020-07-14T22:04:17Z)

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.