# DCoPilot: Generative AI-Empowered Policy Adaptation for Dynamic Data Center Operations

arXiv: https://arxiv.org/abs/2602.02137 · HTML: https://mlsanity.com/paper/2602.02137/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 23 of its 50 references that are in the map.
- First posted on arXiv 2 Feb 2026.

## Builds on (most cited first, 23 of 23)

- [Denoising Diffusion Probabilistic Models](https://mlsanity.com/paper/2006.11239/) (2020)
- [Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks](https://mlsanity.com/paper/1703.03400/) (2017)
- [Diffusion Policy: Visuomotor Policy Learning via Action Diffusion](https://mlsanity.com/paper/2303.04137/) (2023)
- [Progressive Neural Networks](https://mlsanity.com/paper/1606.04671/) (2016)
- [HyperNetworks](https://mlsanity.com/paper/1609.09106/) (2016)
- [AlphaEvolve: A coding agent for scientific and algorithmic discovery](https://mlsanity.com/paper/2506.13131/) (2025)
- [When to Trust Your Model: Model-Based Policy Optimization](https://mlsanity.com/paper/1906.08253/) (2019)
- [Eureka: Human-Level Reward Design via Coding Large Language Models](https://mlsanity.com/paper/2310.12931/) (2023)
- [Efficient Off-Policy Meta-Reinforcement Learning via Probabilistic Context Variables](https://mlsanity.com/paper/1903.08254/) (2019)
- [DrEureka: Language Model Guided Sim-To-Real Transfer](https://mlsanity.com/paper/2406.01967/) (2024)
- [Self-Refined Large Language Model as Automated Reward Function Designer for Deep Reinforcement Learning in Robotics](https://mlsanity.com/paper/2309.06687/) (2023)
- [Recomposing the Reinforcement Learning Building Blocks with Hypernetworks](https://mlsanity.com/paper/2106.06842/) (2021)
- [LLM4PLC: Harnessing Large Language Models for Verifiable Programming of PLCs in Industrial Control Systems](https://mlsanity.com/paper/2401.05443/) (2024)
- [Hypernetworks for Zero-shot Transfer in Reinforcement Learning](https://mlsanity.com/paper/2211.15457/) (2022)
- [Transforming Cooling Optimization for Green Data Center via Deep Reinforcement Learning](https://mlsanity.com/paper/1709.05077/) (2017)
- [Reinforcement Learning Testbed for Power-Consumption Optimization](https://mlsanity.com/paper/1808.10427/) (2018)
- [Pre-Trained Large Language Models for Industrial Control](https://mlsanity.com/paper/2308.03028/) (2023)
- [Deep Reinforcement Learning (DRL): Another Perspective for Unsupervised Wireless Localization](https://mlsanity.com/paper/2004.04618/) (2020)
- [Meta-Controller: Few-Shot Imitation of Unseen Embodiments and Tasks in Continuous Control](https://mlsanity.com/paper/2412.12147/) (2024)
- [Data Center Cooling System Optimization Using Offline Reinforcement Learning](https://mlsanity.com/paper/2501.15085/) (2025)
- [MOSAIC: A Multi-Objective Optimization Framework for Sustainable Datacenter Management](https://mlsanity.com/paper/2311.08583/) (2023)
- [Transforming Future Data Center Operations and Management via Physical AI](https://mlsanity.com/paper/2504.04982/) (2025)
- [Toward Multiphysics-Informed Machine Learning for Sustainable Data Center Operations: Intelligence Evolution with Deployable Solutions for Computing Infrastructure](https://mlsanity.com/paper/2505.19414/) (2025)

## Built on by (most cited first, 1 of 1)

- [Beyond Scripted Search: Sample-Efficient Reward Discovery via Agentic Black-box Optimization](https://mlsanity.com/paper/2609.32394/) (2026)

## Abstract (arXiv)

Modern data centers (DCs) hosting artificial intelligence (AI)-dedicated devices operate at high power densities with rapidly varying workloads, making minute-level adaptation essential for safe and energy-efficient operation. However, manually designing piecewise deep reinforcement learning (DRL) agents cannot keep pace with frequent dynamics shifts and service-level agreement (SLA) changes of an evolving DC. This specification-to-policy lag causes a lack of timely, effective control policies, which may lead to service outages. To bridge the gap, we present DCoPilot, a hybrid framework for generative control policies in dynamic DC operation. DCoPilot synergizes two distinct generative paradigms, i.e., a large language model (LLM) that performs symbolic generation of structured reward forms, and a hypernetwork that conducts parametric generation of policy weights. DCoPilot operates through three coordinated phases: (i) simulation scale-up, which stress-tests reward candidates across diverse simulation-ready (SimReady) scenes; (ii) meta policy distillation, where a hypernetwork is trained to output policy weights conditioned on SLA and scene embeddings; and (iii) online adaptation, enabling zero-shot policy generation in response to updated specifications. Evaluated across five control task families spanning diverse DC components, DCoPilot achieves near-zero constraint violations and outperforms all baselines across specification variations. Ablation studies validate the effectiveness of LLM-based unified reward generation in enabling stable hypernetwork convergence.