# Double Horizon Model-Based Policy Optimization

arXiv: https://arxiv.org/abs/2512.15439 · HTML: https://mlsanity.com/paper/2512.15439/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in RL theory and sample complexity, one of 15 directions in Deep reinforcement learning methods.
- Builds on 24 of its 40 references that are in the map.
- First posted on arXiv 17 Dec 2025.

## Placed in

- Deep reinforcement learning methods / [RL theory and sample complexity](https://mlsanity.com/deeprl/rl-theory-and-sample-complexity/)

## Builds on (most cited first, 24 of 24)

- [Layer Normalization](https://mlsanity.com/paper/1607.06450/) (2016)
- [Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor](https://mlsanity.com/paper/1801.01290/) (2018)
- [Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translation](https://mlsanity.com/paper/1406.1078/) (2014)
- [Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles](https://mlsanity.com/paper/1612.01474/) (2016)
- [Dream to Control: Learning Behaviors by Latent Imagination](https://mlsanity.com/paper/1912.01603/) (2019)
- [Mastering Diverse Domains through World Models](https://mlsanity.com/paper/2301.04104/) (2023)
- [Sigmoid-Weighted Linear Units for Neural Network Function Approximation in Reinforcement Learning](https://mlsanity.com/paper/1702.03118/) (2017)
- [Deep Reinforcement Learning at the Edge of the Statistical Precipice](https://mlsanity.com/paper/2108.13264/) (2021)
- [Deep Reinforcement Learning in a Handful of Trials using Probabilistic Dynamics Models](https://mlsanity.com/paper/1805.12114/) (2018)
- [When to Trust Your Model: Model-Based Policy Optimization](https://mlsanity.com/paper/1906.08253/) (2019)
- [TD-MPC2: Scalable, Robust World Models for Continuous Control](https://mlsanity.com/paper/2310.16828/) (2023)
- [Temporal Difference Learning for Model Predictive Control](https://mlsanity.com/paper/2203.04955/) (2022)
- [dm_control: Software and Tasks for Continuous Control](https://mlsanity.com/paper/2006.12983/) (2020)
- [Randomized Ensembled Double Q-Learning: Learning Fast Without a Model](https://mlsanity.com/paper/2101.05982/) (2021)
- [A Theory of Regularized Markov Decision Processes](https://mlsanity.com/paper/1901.11275/) (2019)
- [Sample-Efficient Reinforcement Learning with Stochastic Ensemble Value Expansion](https://mlsanity.com/paper/1807.01675/) (2018)
- [Learning Continuous Control Policies by Stochastic Value Gradients](https://mlsanity.com/paper/1510.09142/) (2015)
- [Model-Based Value Estimation for Efficient Model-Free Reinforcement Learning](https://mlsanity.com/paper/1803.00101/) (2018)
- [Dropout Q-Functions for Doubly Efficient Reinforcement Learning](https://mlsanity.com/paper/2110.02034/) (2021)
- [PIPPS: Flexible Model-Based Policy Search Robust to the Curse of Chaos](https://mlsanity.com/paper/1902.01240/) (2019)
- [Consistent On-Line Off-Policy Evaluation](https://mlsanity.com/paper/1702.07121/) (2017)
- [On the model-based stochastic value gradient for continuous reinforcement learning](https://mlsanity.com/paper/2008.12775/) (2020)
- [MBRL-Lib: A Modular Library for Model-based Reinforcement Learning](https://mlsanity.com/paper/2104.10159/) (2021)
- [Trust the Model Where It Trusts Itself -- Model-Based Actor-Critic with Uncertainty-Aware Rollout Adaption](https://mlsanity.com/paper/2405.19014/) (2024)

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

- [A Single Deep Preference-Conditioned Policy for Learning Pareto Coverage Sets](https://mlsanity.com/paper/2605.08946/) (2026)

## Abstract (arXiv)

Model-based reinforcement learning (MBRL) reduces the cost of real-environment sampling by generating synthetic trajectories (called rollouts) from a learned dynamics model. However, choosing the length of the rollouts poses two dilemmas: (1) Longer rollouts better preserve on-policy training but amplify model bias, indicating the need for an intermediate horizon to mitigate distribution shift (i.e., the gap between on-policy and past off-policy samples). (2) Moreover, a longer model rollout may reduce value estimation bias but raise the variance of policy gradients due to backpropagation through multiple steps, implying another intermediate horizon for stable gradient estimates. However, these two optimal horizons may differ. To resolve this conflict, we propose Double Horizon Model-Based Policy Optimization (DHMBPO), which divides the rollout procedure into a long "distribution rollout" (DR) and a short "training rollout" (TR). The DR generates on-policy state samples for mitigating distribution shift. In contrast, the short TR leverages differentiable transitions to offer accurate value gradient estimation with stable gradient updates, thereby requiring fewer updates and reducing overall runtime. We demonstrate that the double-horizon approach effectively balances distribution shift, model bias, and gradient instability, and surpasses existing MBRL methods on continuous-control benchmarks in terms of both sample efficiency and runtime.