Research map / Reinforcement learning / Imitation learning

AutoCGP: Closed-Loop Concept-Guided Policies from Unlabeled Demonstrations

ICLR 2025 · Reinforcement learning › Imitation learning · approach: Offline imitation from imperfect data

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes AutoCGP, an imitation learning framework that discovers manipulation concepts from unlabeled demonstrations and trains closed-loop concept-guided robot policies for long-horizon tasks.

Problem

Training embodied agents for complex long-horizon robotic manipulation is hard due to task compositionality, environmental diversity, and dynamic changes, and existing methods often rely on predefined or labeled structure.

Approach

Discovers manipulation concepts automatically from unlabeled demonstrations and uses them to guide closed-loop policies, rather than relying on pre-labeled or hand-specified concepts.

Experimental setup

Datasets: unlabeled robot demonstrations

Closest papers by meaning

• Behavioral Exploration: Learning to Explore via In-Context Adaptation (ICML 2025 · similarity 0.89)
• OSVI-WM: One-Shot Visual Imitation for Unseen Tasks using World-Model-Guided Trajectory Generation (NeurIPS 2025 · similarity 0.89)
• Generalizable Domain Adaptation for Sim-and-Real Policy Co-Training (NeurIPS 2025 · similarity 0.88)
• Chain-of-Action: Trajectory Autoregressive Modeling for Robotic Manipulation (NeurIPS 2025 · similarity 0.88)
• InfoCon: Concept Discovery with Generative and Discriminative Informativeness (ICLR 2024 · similarity 0.88)
• Meta-Controller: Few-Shot Imitation of Unseen Embodiments and Tasks in Continuous Control (NeurIPS 2024 · similarity 0.88)