Online Dynamic Goal Recognition in Gym Environments

Shamir Matan, Elhadad Osher, Nageris Ben, Mirsky Reuth
9/27/2025
cs.ROcs.AI

Official code: MatanShamir1/gr_envs (2 stars, MIT, last commit 2025-09-20). 1 more implementation below.

Abstract

Goal Recognition (GR) is the task of inferring an agent's intended goal from partial observations of its behavior, typically in an online and one-shot setting. Despite recent advances in model-free GR, particularly in applications such as human-robot interaction, surveillance, and assistive systems, the field remains fragmented due to inconsistencies in benchmarks, domains, and evaluation protocols. To address this, we introduce gr-libs (https://github.com/MatanShamir1/gr_libs) and gr-envs (https://github.com/MatanShamir1/gr_envs), two complementary open-source frameworks that support the development, evaluation, and comparison of GR algorithms in Gym-compatible environments. gr-libs includes modular implementations of MDP-based GR baselines, diagnostic tools, and evaluation utilities. gr-envs provides a curated suite of environments adapted for dynamic and goal-directed behavior, along with wrappers that ensure compatibility with standard reinforcement learning toolkits. Together, these libraries offer a standardized, extensible, and reproducible platform for advancing GR research. Both packages are open-source and available on GitHub and PyPI.

View on arXivView PDF

Code Implementations(2)

MatanShamir1/gr_envsOfficial100%
MIT

Gym Environments adjusted to Goal Recognition tasks

20PythonMar 13, 20251 years agoMIT
MatanShamir1/gr_libsOfficial100%
MIT

GR as Metric Learning

21Python, DockerfileMar 2, 20249 months agoMIT

Similar Papers

NiceWebRL: a Python library for human subject experiments with reinforcement learning environments

84% match1 repo842025

MA-VLCM: A Vision Language Critic Model for Value Estimation of Policies in Multi-Agent Team Settings

82% match2026

AHA: A Vision-Language-Model for Detecting and Reasoning Over Failures in Robotic Manipulation

82% match2026

RLInspect: An Interactive Visual Approach to Assess Reinforcement Learning Algorithm

82% match2024

A Vision-Language-Action-Critic Model for Robotic Real-World Reinforcement Learning

81% match1 repo3172025

Real-World Offline Reinforcement Learning from Vision Language Model Feedback

81% match1 repo2024

Cite this paper

@article{matan2025online,
  title  = {Online Dynamic Goal Recognition in Gym Environments},
  author = {Shamir Matan and Elhadad Osher and Nageris Ben and Mirsky Reuth},
  year   = {2025},
  eprint = {2509.23244},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2509.23244v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2509.23244.svg)](https://codeofpaper.com/paper/2509.23244)

Discussion