GRAIL: Goal Recognition Alignment through Imitation Learning

2026-02-15

AI Summary

Traditional goal recognition approaches struggle with suboptimal, biased, or noisy behaviors due to their reliance on idealized assumptions about expert strategies. This work proposes the first goal recognition framework that integrates imitation learning with inverse reinforcement learning: for each candidate goal, it learns a corresponding policy from potentially suboptimal demonstrations and scores observed trajectories via a single forward pass, enabling efficient recognition. The method maintains one-shot inference capability while significantly improving robustness—achieving F1 score gains exceeding 0.5 under systematic bias, improvements of 0.1–0.3 with suboptimal behavior, and up to 0.4 on noisy near-optimal trajectories—while remaining competitive even in fully optimal conditions.

Technology Category

• Machine Learning: Imitation Learning & Inverse Reinforcement Learning
• Intelligent Robots: Learning & Optimization for ROB
• Planning, Routing, and Scheduling: Activity and Plan Recognition

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

Understanding an agent's goals from its behavior is fundamental to aligning AI systems with human intentions. Existing goal recognition methods typically rely on an optimal goal-oriented policy representation, which may differ from the actor's true behavior and hinder the accurate recognition of their goal. To address this gap, this paper introduces Goal Recognition Alignment through Imitation Learning (GRAIL), which leverages imitation learning and inverse reinforcement learning to learn one goal-directed policy for each candidate goal directly from (potentially suboptimal) demonstration trajectories. By scoring an observed partial trajectory with each learned goal-directed policy in a single forward pass, GRAIL retains the one-shot inference capability of classical goal recognition while leveraging learned policies that can capture suboptimal and systematically biased behavior. Across the evaluated domains, GRAIL increases the F1-score by more than 0.5 under systematically biased optimal behavior, achieves gains of approximately 0.1-0.3 under suboptimal behavior, and yields improvements of up to 0.4 under noisy optimal trajectories, while remaining competitive in fully optimal settings. This work contributes toward scalable and robust models for interpreting agent goals in uncertain environments.

Problem

• goal recognition
• suboptimal behavior
• systematic bias
• noisy trajectories
• policy representation

Innovation

• Goal Recognition
• Imitation Learning
• Inverse Reinforcement Learning
• Suboptimal Behavior
• Policy Learning