{
  "abstract": "Bayesian neural networks (BNNs) offer a principled framework for quantifying epistemic uncertainty. However, standard weight-space priors yield complex, high-dimensional neural network posteriors that are computationally prohibitive to sample from, and popular approximate inference schemes often fail to capture epistemic uncertainty effectively---a failure mode particularly acute in reinforcement learning (RL) \\citep{osband2018randomized}. To resolve this challenge, we introduce a new class of BNNs that quantify uncertainty by placing Bayesian nonparametric Dirichlet Process (DP) priors directly on the underlying data-generating distribution rather than on network parameters. Building on this foundation, we develop DP-DQN, a reinforcement learning algorithm that leverages DP-BNNs to maintain posterior distributions over value functions and perform efficient Thompson sampling. By grounding randomized value functions in Bayesian nonparametrics, DP-DQN unifies data-driven posterior reweighting (via the Bayesian Bootstrap) with generative prior hypotheses injected directly through the DP base measure, enabling decisive ``deep exploration'' with a single neural network without the overhead of multi-network ensembles.",
  "abstract_shown": true,
  "affiliations": [],
  "article_type": "Journal reference: Transactions on Machine Learning Research (2026)",
  "arxiv": "2610.27570",
  "author_total": 1,
  "authors": [
    {
      "name": "Sumit Vashishtha"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/cwg95v712qptg.bib",
  "cited_by": 0,
  "doi": "10.48550/arxiv.2610.27570",
  "fulltext_url": null,
  "id": "cwg95v712qptg",
  "issns": [],
  "landing_url": "https://arxiv.org/abs/2610.27570",
  "language": null,
  "licence": "CC-BY-4.0",
  "licence_label": "CC BY 4.0",
  "licence_source": "article",
  "licence_url": "http://creativecommons.org/licenses/by/4.0/",
  "metadata_licence": "CC0-1.0",
  "native_subjects": [
    [
      "arxiv",
      "cs.LG"
    ],
    [
      "arxiv",
      "stat.CO"
    ],
    [
      "arxiv",
      "stat.ML"
    ],
    [
      "openalex",
      "1702|Reinforcement Learning in Robotics"
    ]
  ],
  "openalex": "W7223141340",
  "pdf_url": "https://arxiv.org/pdf/2610.27570",
  "pmcid": null,
  "pmid": null,
  "published": "2026-10-20",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/cwg95v712qptg.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:34:13Z",
      "native_id": "2610.27570",
      "source": "arxiv",
      "url": "https://arxiv.org/abs/2610.27570"
    }
  ],
  "subjects": [
    {
      "field": "mathematics",
      "origin": "arxiv:stat.CO",
      "subfield": "statistics"
    }
  ],
  "title": "Bayesian Neural Networks via Dirichlet Process Priors on the Data-Generating Distribution for Randomized Value Functions in Deep Reinforcement Learning",
  "title_text": "Bayesian Neural Networks via Dirichlet Process Priors on the Data-Generating Distribution for Randomized Value Functions in Deep Reinforcement Learning",
  "type": "preprint",
  "updated_at": "2026-10-30T12:34:13Z",
  "url": "https://research.draconitic.com/paper/cwg95v712qptg",
  "venue": "arXiv",
  "versions": []
}