Think Deep and Fast: Learning Neural Nonlinear Opinion Dynamics from Inverse Dynamic Games for Split-Second Interactions

2024-06-14

arXiv.org

AI Summary

In multi-agent real-time adversarial scenarios (e.g., autonomous racing), fixed-parameter assumptions in opinion dynamics models lead to decision latency and deadlock. Method: We propose Neural Nonlinear Opinion Dynamics (Neural NOD), the first framework embedding neural differential equations into an inverse dynamic game formulation to jointly learn time-varying opinion evolution parameters end-to-end from expert trajectories—enabling game-theoretic, adaptive modeling. We further introduce a multi-agent joint state-action representation to enhance cooperative perception. Contribution/Results: Evaluated on simulated racing tasks, Neural NOD achieves millisecond-scale response times, significantly reduces collision rates, and improves overtaking success by 12.7% over fixed-parameter NOD and state-of-the-art inverse-game baselines. It establishes a novel, learnable, and interpretable paradigm for rapid, deadlock-free decision-making in dynamic adversarial environments.

Technology Category

• Multiagent Systems: Adversarial Agents
• Game Theory and Economic Paradigms: Adversarial Learning
• Search and Optimization: Adversarial Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Non-cooperative interactions commonly occur in multi-agent scenarios such as car racing, where an ego vehicle can choose to overtake the rival, or stay behind it until a safe overtaking"corridor"opens. While an expert human can do well at making such time-sensitive decisions, autonomous agents are incapable of rapidly reasoning about complex, potentially conflicting options, leading to suboptimal behaviors such as deadlocks. Recently, the nonlinear opinion dynamics (NOD) model has proven to exhibit fast opinion formation and avoidance of decision deadlocks. However, NOD modeling parameters are oftentimes assumed fixed, limiting their applicability in complex and dynamic environments. It remains an open challenge to determine such parameters automatically and adaptively, accounting for the ever-changing environment. In this work, we propose for the first time a learning-based and game-theoretic approach to synthesize a Neural NOD model from expert demonstrations, given as a dataset containing (possibly incomplete) state and action trajectories of interacting agents. We demonstrate Neural NOD's ability to make fast and deadlock-free decisions in a simulated autonomous racing example. We find that Neural NOD consistently outperforms the state-of-the-art data-driven inverse game baseline in terms of safety and overtaking performance.

Problem

• Autonomous agents struggle with fast, complex decision-making in interactions
• Fixed NOD parameters limit adaptability in dynamic environments
• Need for adaptive parameter learning from expert demonstrations

Innovation

• Learning Neural NOD from inverse dynamic games
• Adaptive parameter determination for dynamic environments
• Fast deadlock-free decisions in autonomous racing