mlsanityData to 13 Oct 2026

Explaining How a Neural Network Play the Go Game and Let People Learn

arXiv 2310.09838 ↗First posted 15 Oct 2023cs.AI cs.CV cs.LGcited by 6

Not on the map. Why?

Abstract (arXiv)

The AI model has surpassed human players in the game of Go, and it is widely believed that the AI model has encoded new knowledge about the Go game beyond human players. In this way, explaining the knowledge encoded by the AI model and using it to teach human players represent a promising-yet-challenging issue in explainable AI. To this end, mathematical supports are required to ensure that human players can learn accurate and verifiable knowledge, rather than specious intuitive analysis. Thus, in this paper, we extract interaction primitives between stones encoded by the value network for the Go game, so as to enable people to learn from the value network. Experiments show the effectiveness of our method.

Key facts. Cited by 6 papers in the map (data through 13 Oct 2026). Builds on 14 of its 24 references that are in the map. First posted on arXiv 15 Oct 2023.

Builds on

14 of 24 references.

References · 14

  1. 2016Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localizationcited by 6,713
  2. 2017A Unified Approach to Interpreting Model Predictionscited by 6,515
  3. 2015Learning Deep Features for Discriminative Localizationcited by 2,434
  4. 2013Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Mapscited by 2,360
  5. 2013Visualizing and Understanding Convolutional Networkscited by 2,282
  6. 2015Understanding Neural Networks Through Deep Visualizationcited by 333
  7. 2015Inverting Visual Representations with Convolutional Networkscited by 168
  8. 2017Visualizing Deep Neural Network Decisions: Prediction Difference Analysiscited by 166
  9. 2019The Shapley Taylor Interaction Indexcited by 143
  10. 2019Accelerating Self-Play Learning in Gocited by 61
  11. 2021Defining and Quantifying the Emergence of Sparse Concepts in DNNscited by 36
  12. 2023Does a Neural Network Really Encode Symbolic Concepts?cited by 27
2 more
  1. 2023Where We Have Arrived in Proving the Emergence of Sparse Symbolic Concepts in AI Modelscited by 8
  2. 2023Towards Attributions of Input Variables in a Coalitioncited by 1

Built on by

Cited by year

Citing papers per year: 2024: 6.

6 papers.

  1. 2024Defining and Extracting generalizable interaction primitives from DNNscited by 14
  2. 2024Towards the Dynamics of a DNN Learning Symbolic Interactionscited by 13
  3. 2024Two-Phase Dynamics of Interactions Explains the Starting Point of a DNN Learning Over-Fitted Featurescited by 5
  4. 2024Layerwise Change of Knowledge in Neural Networkscited by 4
  5. 2024Quantifying In-Context Reasoning Effects and Memorization Effects in LLMscited by 3
  6. 2024Disentangling Regional Primitives for Image Generationcited by 0