PolicyBank: Evolving Policy Understanding for LLM Agents

• URL: http://arxiv.org/abs/2604.15505v1
• Date: Thu, 16 Apr 2026 20:29:30 GMT
• Title: PolicyBank: Evolving Policy Understanding for LLM Agents
• Abstract summary: PolicyBank maintains structured, tool-level policy insights and iteratively refines them.<n>PolicyBank closes up to 82% of the gap toward a human oracle.
• Score: 51.86716874651299
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: LLM agents operating under organizational policies must comply with authorization constraints typically specified in natural language. In practice, such specifications inevitably contain ambiguities and logical or semantic gaps that cause the agent's behavior to systematically diverge from the true requirements. We ask: by letting an agent evolve its policy understanding through interaction and corrective feedback from pre-deployment testing, can it autonomously refine its interpretation to close specification gaps? We propose PolicyBank, a memory mechanism that maintains structured, tool-level policy insights and iteratively refines them -- unlike existing memory mechanisms that treat the policy as immutable ground truth, reinforcing "compliant but wrong" behaviors. We also contribute a systematic testbed by extending a popular tool-calling benchmark with controlled policy gaps that isolate alignment failures from execution failures. While existing memory mechanisms achieve near-zero success on policy-gap scenarios, PolicyBank closes up to 82% of the gap toward a human oracle.

Related papers

• PolicyMem: Geometric Policy Memory for LLM Governance [72.04962480384322]
  We introduce PolicyMem, a geometric policy memory that externalizes natural-language policies as reusable geometric memory objects.<n>A memory writer compiles natural-language policies into policy memory slots, and query-response pairs read the policy memory through projection energy.<n>The resulting policy-evidence profile directly mediates the safety verdict and is reused for policy attribution and post-intervention verification.
  arXiv  Detail & Related papers  (2026-09-12T06:07:43Z)
• EMERGE-Policy: A Robot Mind Emerges Beyond a Single Policy [48.898967774856466]
  EMERGE-Policy coordinates both capability invocation and information exchange.<n>Main Agent retains task-level state within an active context window.<n>Role-specific Sub Agents process perception, execution monitoring, verification, and memory consolidation.
  arXiv  Detail & Related papers  (2026-08-30T16:46:06Z)
• OGPO: Sample Efficient Full-Finetuning of Generative Control Policies [53.42266064673132]
  Generative control policies (GCPs) have emerged as effective parameterizations for robot learning.<n>This work introduces Off-policy Generative Policy Optimization (OGPO), a sample-efficient for finetuning GCPs.<n>OGPO achieves state-of-the-art performance on manipulation tasks spanning multi-task settings, high-precision insertion, and dexterous control.
  arXiv  Detail & Related papers  (2026-05-04T18:36:40Z)
• Near-Miss: Latent Policy Failure Detection in Agentic Workflows [9.719140082591956]
  We introduce a novel metric for detecting latent policy failures in agent conversations traces.<n>Our results show that latent failures occur in 8-17% of trajectories involving mutating tool calls.
  arXiv  Detail & Related papers  (2026-03-31T12:26:35Z)
• When to Act, Ask, or Learn: Uncertainty-Aware Policy Steering [10.01278648231868]
  Policy steering is an emerging way to adapt robot behaviors at deployment-time.<n> Vision-Language Models (VLMs) are promising general-purpose verifiers due to their reasoning capabilities.<n>We propose uncertainty-aware policy steering (UPS), a framework that jointly reasons about semantic task uncertainty and low-level action feasibility.
  arXiv  Detail & Related papers  (2026-02-25T23:23:22Z)
• Training-Free Policy Violation Detection via Activation-Space Whitening in LLMs [21.5603664964501]
  We propose a training-free and efficient method that treats policy violation detection as an out-of-distribution detection problem.<n>Inspired by whitening techniques, we apply a linear transformation to decorrelate the model's hidden activations and standardize them to zero mean and unit variance.<n>On a challenging policy benchmark, our approach achieves state-of-the-art results, surpassing both existing guardrails and fine-tuned reasoning models.
  arXiv  Detail & Related papers  (2025-12-03T17:23:39Z)
• Learning Deterministic Policies with Policy Gradients in Constrained Markov Decision Processes [59.27926064817273]
  We introduce an exploration-agnostic algorithm, called C-PG, which enjoys global last-iterate convergence guarantees under domination assumptions.<n>We empirically validate both the action-based (C-PGAE) and parameter-based (C-PGPE) variants of C-PG on constrained control tasks.
  arXiv  Detail & Related papers  (2025-06-06T10:29:05Z)
• PARL: A Unified Framework for Policy Alignment in Reinforcement Learning from Human Feedback [106.63518036538163]
  We present a novel unified bilevel optimization-based framework, textsfPARL, formulated to address the recently highlighted critical issue of policy alignment in reinforcement learning. Our framework addressed these concerns by explicitly parameterizing the distribution of the upper alignment objective (reward design) by the lower optimal variable. Our empirical results substantiate that the proposed textsfPARL can address the alignment concerns in RL by showing significant improvements.
  arXiv  Detail & Related papers  (2023-08-03T18:03:44Z)
• Policy Dispersion in Non-Markovian Environment [53.05904889617441]
  This paper tries to learn the diverse policies from the history of state-action pairs under a non-Markovian environment. We first adopt a transformer-based method to learn policy embeddings. Then, we stack the policy embeddings to construct a dispersion matrix to induce a set of diverse policies.
  arXiv  Detail & Related papers  (2023-02-28T11:58:39Z)
• Bounded Robustness in Reinforcement Learning via Lexicographic Objectives [54.00072722686121]
  Policy robustness in Reinforcement Learning may not be desirable at any cost. We study how policies can be maximally robust to arbitrary observational noise. We propose a robustness-inducing scheme, applicable to any policy algorithm, that trades off expected policy utility for robustness.
  arXiv  Detail & Related papers  (2022-09-30T08:53:18Z)
• Policy Regularization for Legible Behavior [0.0]
  In Reinforcement Learning interpretability generally means to provide insight into the agent's mechanisms. This paper borrows from the Explainable Planning literature methods that focus on the legibility of the agent. In our formulation, the decision boundary introduced by legibility impacts the states in which the agent's policy returns an action that has high likelihood also in other policies.
  arXiv  Detail & Related papers  (2022-03-08T10:55:46Z)
• Learning Goal-oriented Dialogue Policy with Opposite Agent Awareness [116.804536884437]
  We propose an opposite behavior aware framework for policy learning in goal-oriented dialogues. We estimate the opposite agent's policy from its behavior and use this estimation to improve the target agent by regarding it as part of the target policy.
  arXiv  Detail & Related papers  (2020-04-21T03:13:44Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.