aiqbench analysis
This paper introduces GLiBRL, a deep Bayesian reinforcement learning framework that eliminates variational inference errors by using conjugate Bayesian inference with generalized linear task models. By enabling exact sequential posterior updates and permutation-invariant representations, the method achieves superior zero-shot performance on standard locomotion and manipulation benchmarks compared to eight recent meta-RL approaches. The work advances the theoretical robustness of adaptive agents in unseen environments while maintaining compatibility with both off- and on-policy algorithms.
Meta-RL with Bayesian Linear Task Models
arXiv:2512.20974v4 Announce Type: replace-cross Abstract: Deep Bayesian reinforcement learning adapts to unseen tasks by inferring latent transition and reward models, but existing methods typically rely on variational posteriors and evidence lower bounds, introducing approximation error and unstable task representations. We introduce GLiBRL, a deep Bayesian RL framework that combines generalised linear task models with learnable non-linear basis functions. GLiBRL features conjugate Bayesian inference, yielding exact, sequential posterior updates over task parameters and model noise, together with a closed-form marginal likelihood that eliminates variational inference. The update is naturally permutation-invariant, allowing GLiBRL to integrate with both off- and on-policy algorithms. GLiBRL also learns task representation admitting an exact kernel identity, relating distances between task representations to kernel discrepancies over the task contexts. Compared against eight representative or recent meta reinforcement learning methods, GLiBRL achieves the highest aggregate zero-shot test performance on both the MuJoCo locomotion and MetaWorld manipulation benchmarks.
Abstract:Deep Bayesian reinforcement learning adapts to unseen tasks by inferring latent transition and reward models, but existing methods typically rely on variational posteriors and evidence lower bounds, introducing approximation error and unstable task representations. We introduce GLiBRL, a deep Bayesian RL framework that combines generalised linear task models with learnable non-linear basis functions. GLiBRL features conjugate Bayesian inference, yielding exact, sequential posterior updates over task parameters and model noise, together with a closed-form marginal likelihood that eliminates variational inference. The update is naturally permutation-invariant, allowing GLiBRL to integrate with both off- and on-policy algorithms. GLiBRL also learns task representation admitting an exact kernel identity, relating distances between task representations to kernel discrepancies over the task contexts. Compared against eight representative or recent meta reinforcement learning methods, GLiBRL achieves the highest aggregate zero-shot test performance on both the MuJoCo locomotion and MetaWorld manipulation benchmarks.
Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Robotics (cs.RO)
arXiv:2512.20974 [cs.LG]
(or arXiv:2512.20974v4 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2512.20974 Focus to learn more arXiv-issued DOI via DataCite
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 · 240 words
Related coverage
MetaKoopman: Bayesian Meta-Learning of Koopman Operators for Modeling Structured Dynamics under Distribution Shifts
arXiv · Jul 30, 2026
Few-Shot Bioactivity Prediction with Meta-Learning under Assay Heterogeneity
arXiv · Oct 7, 2026
Generalization and infinite width
LessWrong · Aug 1, 2026
[AINews] Meta Connect 2026: Muse glasses, voice, video, and Charm
Latent Space · Sep 24, 2026