What question did this study set out to answer?

The aim is to address semantic alignment in robotic tasks using a neuro-symbolic framework that integrates planning and verification.

Neuro-symbolic Hierarchical Learning for Long-Horizon Robotic Tasks

View Full Paper

Ask AI

Bookmark

Authors

Randomized trial demonstrates improved task execution in robotic systems, suggesting effective integration of symbolic reasoning and control.

The aim is to address semantic alignment in robotic tasks using a neuro-symbolic framework that integrates planning and verification.
Developed a closed-loop, counterexample-guided synthesis framework unifying LLM-based planning and formal verification.
Converted natural language task descriptions into PDDL and generated high-level plans through a Guess-Check-Critique loop.
Co-optimized symbolic abstractions with low-level policies in behavior trees for robustness and executability.
Achieved superior performance on long-horizon tasks compared to state-of-the-art methods.
Provided improved interpretability of robotic executions and enhanced generalization capabilities.

Huang et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark

Randomized trial demonstrates improved task execution in robotic systems, suggesting effective integration of symbolic reasoning and control.