Compliance2LoRA: Personalizable On-Demand Safety Alignment on Arbitrary Policy Subsets via Hypernetwork-Generated LoRA Adapters

• URL: http://arxiv.org/abs/2607.27594v3
• Date: Wed, 05 Aug 2026 18:01:34 GMT
• Title: Compliance2LoRA: Personalizable On-Demand Safety Alignment on Arbitrary Policy Subsets via Hypernetwork-Generated LoRA Adapters
• Abstract summary: Post-training alignment in large reasoning models (LRMs) has significantly improved their adaptability to diverse safety compliance settings.<n>As LRMs personalization for downstream users takes center stage, the demand for varying levels of policy compliance grows.<n>We propose a unified adaptive hypernetwork-based framework for multi-policy compliance.
• Score: 50.29667251847595
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Post-training alignment in large reasoning models (LRMs) has significantly improved their adaptability to diverse safety compliance settings. However, as LRMs personalization for downstream users takes center stage, the demand for varying levels of policy compliance grows as different user-specific LRMs must adhere to distinct subsets of safety policies. Training a separate LRM for each policy subset introduces severe combinatorial overhead. While in context learning methods overcome this combinatorial overhead, they introduce additional computational challenges associated with long context generation. To address this challenge, we propose \ours, a unified adaptive hypernetwork-based framework for multi-policy compliance. In our framework, safety policies serve as customizable inputs to a LoRA adapter generator, which learns to produce policy compliant LoRA weights for downstream LRM. When added to the LRM these weights enable the generation of responses compliant with the specified policy subsets. In this work, we demonstrate that training such a hypernetwork enables on-demand policy adjustments on a single LRM without sacrificing task performance across reasoning models of different sized and different evaluation datasets. This highlights the effectiveness and practicality of adaptive hypernetwork based alignment in LRMs.

Related papers

• Replacing Parameters with Preferences: Federated Alignment of Heterogeneous Vision-Language Models [63.70401095689976]
  Vision-Language Models (VLMs) have broad potential in privacy-sensitive domains such as healthcare and finance.<n>We propose MoR, a federated alignment framework that combines GRPO with Mixture-of-Rewards for heterogeneousVLMs.<n>MoR consistently outperforms federated alignment baselines in generalization and cross-client adaptability.
  arXiv  Detail & Related papers  (2026-05-05T07:02:50Z)
• Safe Multi-Agent Deep Reinforcement Learning for Privacy-Aware Edge-Device Collaborative DNN Inference [8.14391361533752]
  This paper proposes a privacy-aware collaborative inference framework, in which adaptive model partitioning is performed across edge devices and servers.<n>We formulate the joint problem as a Constrained Markov Decision Process (CMDP) that integrates model deployment, user-server association, model partitioning, and resource allocation.<n>We show that HC-MAPPO-L consistently satisfies stringent delay constraints while achieving a superior balance among energy consumption and privacy cost.
  arXiv  Detail & Related papers  (2026-02-23T11:33:52Z)
• MAESTRO: Meta-learning Adaptive Estimation of Scalarization Trade-offs for Reward Optimization [56.074760766965085]
  Group-Relative Policy Optimization has emerged as an efficient paradigm for aligning Large Language Models (LLMs)<n>We propose MAESTRO, which treats reward scalarization as a dynamic latent policy, leveraging the model's terminal hidden states as a semantic bottleneck.<n>We formulate this as a contextual bandit problem within a bi-level optimization framework, where a lightweight Conductor network co-evolves with the policy by utilizing group-relative advantages as a meta-reward signal.
  arXiv  Detail & Related papers  (2026-01-12T05:02:48Z)
• Collab: Controlled Decoding using Mixture of Agents for LLM Alignment [90.6117569025754]
  Reinforcement learning from human feedback has emerged as an effective technique to align Large Language models.<n>Controlled Decoding provides a mechanism for aligning a model at inference time without retraining.<n>We propose a mixture of agent-based decoding strategies leveraging the existing off-the-shelf aligned LLM policies.
  arXiv  Detail & Related papers  (2025-03-27T17:34:25Z)
• AsymLoRA: Harmonizing Data Conflicts and Commonalities in MLLMs [5.018961516699825]
  AsymLoRA is a parameter-efficient tuning framework that unifies knowledge modularization and cross-modal coordination.<n>AsymLoRA consistently surpasses both vanilla LoRA, which captures only commonalities, and LoRA-MoE, which focuses solely on conflicts.
  arXiv  Detail & Related papers  (2025-02-27T12:21:02Z)
• Low-Rank Agent-Specific Adaptation (LoRASA) for Multi-Agent Policy Learning [3.333453555166201]
  Multi-agent reinforcement learning (MARL) often relies on emph parameter sharing (PS) to scale efficiently.<n>We propose textbfLow-Rank Agent-Specific Adaptation (LoRASA), a novel approach that treats each agent's policy as a specialized task'' fine-tuned from a shared backbone.<n>We evaluate LoRASA on challenging benchmarks including the StarCraft Multi-Agent Challenge (SMAC) and Multi-Agent MuJoCo (MAMuJoCo)
  arXiv  Detail & Related papers  (2025-02-08T13:57:53Z)
• Large Language Model driven Policy Exploration for Recommender Systems [50.70228564385797]
  offline RL policies trained on static user data are vulnerable to distribution shift when deployed in dynamic online environments.<n>Online RL-based RS also face challenges in production deployment due to the risks of exposing users to untrained or unstable policies.<n>Large Language Models (LLMs) offer a promising solution to mimic user objectives and preferences for pre-training policies offline.<n>We propose an Interaction-Augmented Learned Policy (iALP) that utilizes user preferences distilled from an LLM.
  arXiv  Detail & Related papers  (2025-01-23T16:37:44Z)
• HarmoDT: Harmony Multi-Task Decision Transformer for Offline Reinforcement Learning [72.25707314772254]
  We introduce the Harmony Multi-Task Decision Transformer (HarmoDT), a novel solution designed to identify an optimal harmony subspace of parameters for each task. The upper level of this framework is dedicated to learning a task-specific mask that delineates the harmony subspace, while the inner level focuses on updating parameters to enhance the overall performance of the unified policy.
  arXiv  Detail & Related papers  (2024-05-28T11:41:41Z)
• Personalized Reinforcement Learning with a Budget of Policies [9.846353643883443]
  Personalization in machine learning (ML) tailors models' decisions to the individual characteristics of users. We propose a novel framework termed represented Markov Decision Processes (r-MDPs) that is designed to balance the need for personalization with the regulatory constraints. In an r-MDP, we cater to a diverse user population, each with unique preferences, through interaction with a small set of representative policies. We develop two deep reinforcement learning algorithms that efficiently solve r-MDPs.
  arXiv  Detail & Related papers  (2024-01-12T11:27:55Z)
• A State-Augmented Approach for Learning Optimal Resource Management Decisions in Wireless Networks [58.720142291102135]
  We consider a radio resource management (RRM) problem in a multi-user wireless network. The goal is to optimize a network-wide utility function subject to constraints on the ergodic average performance of users. We propose a state-augmented parameterization for the RRM policy, where alongside the instantaneous network states, the RRM policy takes as input the set of dual variables corresponding to the constraints.
  arXiv  Detail & Related papers  (2022-10-28T21:24:13Z)

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.