[Submitted on 24 Dec 2025 (v1), last revised 17 Nov 2026 (this version, v5)]
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.
Submission history
From: Jingyang You [view email]
[v1] Wed, 24 Dec 2025 06:00:51 UTC (3,218 KB)
[v2] Thu, 7 May 2026 05:25:01 UTC (4,837 KB)
[v3] Fri, 8 May 2026 01:20:50 UTC (4,834 KB)
[v4] Wed, 9 Sep 2026 09:27:59 UTC (1,420 KB)
[v5] Tue, 17 Nov 2026 23:10:19 UTC (1,299 KB)
