Disentangling the QiGan Encoded by a DNN towards the Go Game

Huilin Zhou, Huijie Tang, Mingjie Li, Hao Zhang, Zhen Yu Liu, Edouardo Honig, Ying Nian Wu, Liang Lin, Lianghui Ding, Quanshi Zhang

Keywords: Katago, Knowledge discovery

TL;DR: This paper extracts shape patterns that explain the QiGan encoded by the value network for the Go game.

Abstract: Given a deep neural network (DNN) that has surpassed human beings in a task, disentangling the explicit knowledge encoded by the DNN to obtain some new insights into the task is a new promising-yet-challenging regime in explainable AI. In this paper, we aim to disentangle the ''QiGan'' encoded by the AI model for the Go game, which has beat top human players. Specifically, we disentangle primitive shape patterns of stones memorized by the value network, and these shape patterns represent the ''QiGan'' used to conduct a fast situation assessment of the current board state. The universal-matching property of interactions ensure that human players can learn accurate and verifiable shape patterns, rather than specious intuitive analysis. In experiments, our method explains lots of novel shape patterns beyond traditional shape patterns in human knowledge.

Primary Area: interpretability and explainable AI

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.

Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.

Reciprocal Reviewing: I understand the reciprocal reviewing requirement as described on https://iclr.cc/Conferences/2025/CallForPapers. If none of the authors are registered as a reviewer, it may result in a desk rejection at the discretion of the program chairs. To request an exception, please complete this form at https://forms.gle/Huojr6VjkFxiQsUp6.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Submission Number: 2649