# Beyond Myopic World Models: Long-Horizon End-to-End Training for Direct Future Prediction

arXiv: https://arxiv.org/abs/2608.07420 · HTML: https://mlsanity.com/paper/2608.07420/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Latent predictive world models, one of 11 directions in World models.
- Builds on 26 of its 43 references that are in the map.
- First posted on arXiv 7 Aug 2026.

## Placed in

- World models / [Latent predictive world models](https://mlsanity.com/wm/latent-predictive-world-models/)

## Builds on (most cited first, 26 of 26)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [RoFormer: Enhanced Transformer with Rotary Position Embedding](https://mlsanity.com/paper/2104.09864/) (2021)
- [FiLM: Visual Reasoning with a General Conditioning Layer](https://mlsanity.com/paper/1709.07871/) (2017)
- [Dream to Control: Learning Behaviors by Latent Imagination](https://mlsanity.com/paper/1912.01603/) (2019)
- [Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware](https://mlsanity.com/paper/2304.13705/) (2023)
- [World Models](https://mlsanity.com/paper/1803.10122/) (2018)
- [Mastering Atari with Discrete World Models](https://mlsanity.com/paper/2010.02193/) (2020)
- [Scheduled Sampling for Sequence Prediction with Recurrent Neural Networks](https://mlsanity.com/paper/1506.03099/) (2015)
- [Diversity is All You Need: Learning Skills without a Reward Function](https://mlsanity.com/paper/1802.06070/) (2018)
- [Genie: Generative Interactive Environments](https://mlsanity.com/paper/2402.15391/) (2024)
- [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)
- [Unsupervised Learning for Physical Interaction through Video Prediction](https://mlsanity.com/paper/1605.07157/) (2016)
- [How Far is Video Generation from World Model: A Physical Law Perspective](https://mlsanity.com/paper/2411.02385/) (2024)
- [Lipschitz Continuity in Model-based Reinforcement Learning](https://mlsanity.com/paper/1804.07193/) (2018)
- [Combating the Compounding-Error Problem with a Multi-step Model](https://mlsanity.com/paper/1905.13320/) (2019)
- [Temporal Abstraction in Reinforcement Learning with the Successor Representation](https://mlsanity.com/paper/2110.05740/) (2021)
- [What Has a Foundation Model Found? Using Inductive Bias to Probe for World Models](https://mlsanity.com/paper/2507.06952/) (2025)
- [Self-Correcting Models for Model-Based Reinforcement Learning](https://mlsanity.com/paper/1612.06018/) (2016)
- [Hierarchical Planning with Latent World Models](https://mlsanity.com/paper/2604.03208/) (2026)
- [Time-Agnostic Prediction: Predicting Predictable Video Frames](https://mlsanity.com/paper/1808.07784/) (2018)
- [Model-Based Planning with Discrete and Continuous Actions](https://mlsanity.com/paper/1705.07177/) (2017)
- [From Kepler to Newton: Inductive Biases Guide Learned World Models in Transformers](https://mlsanity.com/paper/2602.06923/) (2026)
- [Leveraging Jumpy Models for Planning and Fast Learning in Robotic Domains](https://mlsanity.com/paper/2302.12617/) (2023)
- [SmallWorlds: Assessing Dynamics Understanding of World Models in Isolated Environments](https://mlsanity.com/paper/2511.23465/) (2025)
- [Cloning Deterministic Worlds: The Critical Role of Latent Geometry in Long-Horizon World Models](https://mlsanity.com/paper/2510.26782/) (2025)

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

- [Long-Horizon Textual World Modeling through Structured Reasoning](https://mlsanity.com/paper/2610.06637/) (2026)

## Abstract (arXiv)

World models are expected to support imagination over extended temporal horizons, yet most are still trained through local few-step prediction objectives and deployed by recursively rolling out their own predictions. This creates a fundamental mismatch: few-step losses optimize local transition fidelity, while long-horizon prediction depends on how errors and gradients propagate through the entire trajectory. As a result, transitions with different downstream influence on the endpoint are treated uniformly during training, and small local errors are amplified through recursive inference. We argue that long-horizon accuracy is better achieved by optimizing directly, through an end-to-end endpoint prediction objective. To instantiate this paradigm, we introduce the Direct Prediction World Model (DPWM), a non-recursive architecture that compresses an action sequence of arbitrary length into a single embedding and predicts the endpoint observation in a single forward pass. This design avoids recurrent rollout in both prediction and gradient propagation, making long-horizon end-to-end training practical at horizons where unrolled autoregressive training becomes unstable. Empirically, DPWM substantially improves long-horizon endpoint prediction over recursive world-model baselines on continuous-control and pixel-based benchmarks, with larger gains as the prediction horizon increases. We further show that recurrent baselines benefit similarly when retrained with the same long-horizon endpoint objective, supporting our central claim that the training objective, rather than the particular backbone choice, is the main driver of long-horizon prediction accuracy. Our results suggest that world models can benefit from being trained and evaluated at the temporal scales where they are ultimately used, shifting the focus from local transition modeling toward long-horizon predictive accuracy.