Are AlphaZero-like Agents Robust to Adversarial Perturbations?

11/07/2022
by   Li-Cheng Lan, et al.
1

The success of AlphaZero (AZ) has demonstrated that neural-network-based Go AIs can surpass human performance by a large margin. Given that the state space of Go is extremely large and a human player can play the game from any legal state, we ask whether adversarial states exist for Go AIs that may lead them to play surprisingly wrong actions. In this paper, we first extend the concept of adversarial examples to the game of Go: we generate perturbed states that are “semantically” equivalent to the original state by adding meaningless moves to the game, and an adversarial state is a perturbed state leading to an undoubtedly inferior action that is obvious even for Go beginners. However, searching the adversarial state is challenging due to the large, discrete, and non-differentiable search space. To tackle this challenge, we develop the first adversarial attack on Go AIs that can efficiently search for adversarial states by strategically reducing the search space. This method can also be extended to other board games such as NoGo. Experimentally, we show that the actions taken by both Policy-Value neural network (PV-NN) and Monte Carlo tree search (MCTS) can be misled by adding one or two meaningless stones; for example, on 58% of the AlphaGo Zero self-play games, our method can make the widely used KataGo agent with 50 simulations of MCTS plays a losing action by adding two meaningless stones. We additionally evaluated the adversarial examples found by our algorithm with amateur human Go players and 90% of examples indeed lead the Go agent to play an obviously inferior action. Our code is available at <https://PaperCode.cc/GoAttack>.

READ FULL TEXT

page 2

page 10

page 18

research
03/16/2018

Semantic Adversarial Examples

Deep neural networks are known to be vulnerable to adversarial examples,...
research
10/06/2021

No-Press Diplomacy from Scratch

Prior AI successes in complex games have largely focused on settings wit...
research
07/05/2021

Dealing with Adversarial Player Strategies in the Neural Network Game iNNk through Ensemble Learning

Applying neural network (NN) methods in games can lead to various new an...
research
08/25/2022

Turning Mathematics Problems into Games: Reinforcement Learning and Gröbner bases together solve Integer Feasibility Problems

Can agents be trained to answer difficult mathematical questions by play...
research
05/30/2022

Elastic Monte Carlo Tree Search with State Abstraction for Strategy Game Playing

Strategy video games challenge AI agents with their combinatorial search...
research
12/22/2020

Learning to Play Imperfect-Information Games by Imitating an Oracle Planner

We consider learning to play multiplayer imperfect-information games wit...
research
02/17/2018

A Deep Q-Learning Agent for the L-Game with Variable Batch Training

We employ the Deep Q-Learning algorithm with Experience Replay to train ...

Please sign up or login with your details

Forgot password? Click here to reset