aiqbench analysis
This research introduces BEACON, a framework for imitation learning in robotics that resolves state aliasing by conditioning diffusion policies on Bayesian belief representations rather than raw observation history. By structuring uncertainty into both continuous and discrete formats, the method enables agents to dynamically modulate exploration versus exploitation without explicit reward shaping or mode switching. The approach demonstrates significant performance gains over baselines in partially observable manipulation tasks, offering a robust solution for handling hidden states in real-world robotic applications.
BEACON: Belief-Enabled Adaptive CONtrol for Imitation Learning under Uncertainty
arXiv:2609.22730v1 Announce Type: new Abstract: Robot manipulation tasks often involve hidden state information that cannot be directly observed and must be inferred through sequential physical interactions. In such partially observable settings, conditioning an imitation learning policy directly on the recent raw observation history leads to poor performance. This is due to state aliasing, wherein identical observations may arise from different hidden states, and the policy receives conflicting action labels for the same input. To enable history-aware disambiguation capability, we propose conditioning a diffusion policy on a structured representation of the hidden states using Bayesian belief that exposes both the current most likely state estimate and the remaining uncertainty. This representation replaces raw history with a structured, compact input, enabling the policy to implicitly modulate between exploratory and exploitative behaviors based on belief uncertainty, without explicit mode switching or reward shaping. We evaluate across two domains with qualitatively different belief representations: a continuous belief for cornstalk gripper alignment via tactile sensing, and a discrete categorical distribution for latched door opening. In both domains, the belief-conditioned policy substantially outperforms the observation-only baseline and approaches privileged ground-truth performance, with ablations illustrating that the policy adapts its exploration behavior depending on the belief uncertainty at inference time.
Abstract:Robot manipulation tasks often involve hidden state information that cannot be directly observed and must be inferred through sequential physical interactions. In such partially observable settings, conditioning an imitation learning policy directly on the recent raw observation history leads to poor performance. This is due to state aliasing, wherein identical observations may arise from different hidden states, and the policy receives conflicting action labels for the same input. To enable history-aware disambiguation capability, we propose conditioning a diffusion policy on a structured representation of the hidden states using Bayesian belief that exposes both the current most likely state estimate and the remaining uncertainty. This representation replaces raw history with a structured, compact input, enabling the policy to implicitly modulate between exploratory and exploitative behaviors based on belief uncertainty, without explicit mode switching or reward shaping. We evaluate across two domains with qualitatively different belief representations: a continuous belief for cornstalk gripper alignment via tactile sensing, and a discrete categorical distribution for latched door opening. In both domains, the belief-conditioned policy substantially outperforms the observation-only baseline and approaches privileged ground-truth performance, with ablations illustrating that the policy adapts its exploration behavior depending on the belief uncertainty at inference time.
Robotics (cs.RO)
arXiv:2609.22730 [cs.RO]
(or arXiv:2609.22730v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2609.22730 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 291 words
Related coverage
The Machines that Make the Machines
NVIDIA · Oct 7, 2026
Beyond Tech’s Power: On Status, Sacrifice, and the Search for Legitimacy
LessWrong · Aug 4, 2026
The J-Space Debate, Agent Swarms, and Pacing Frontier AI - Digital Minds Newsletter #4
LessWrong · Sep 18, 2026
Robust Belief-State Policy Learning for Quantum Network Routing Under Decoherence and Time-Varying Conditions
arXiv · Jul 23, 2026