# Recommendation World Models for Future-State Control

arXiv: https://arxiv.org/abs/2609.30711 · HTML: https://mlsanity.com/paper/2609.30711/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Latent dynamics models and model-based RL, one of 11 directions in World models.
- Builds on 10 of its 39 references that are in the map.
- First posted on arXiv 25 Sep 2026.

## Placed in

- World models / [Latent dynamics models and model-based RL](https://mlsanity.com/wm/latent-dynamics-models-and-model-based-rl/)

## Builds on (most cited first, 10 of 10)

- [Dream to Control: Learning Behaviors by Latent Imagination](https://mlsanity.com/paper/1912.01603/) (2019)
- [Self-Attentive Sequential Recommendation](https://mlsanity.com/paper/1808.09781/) (2018)
- [BERT4Rec: Sequential Recommendation with Bidirectional Encoder Representations from Transformer](https://mlsanity.com/paper/1904.06690/) (2019)
- [Session-based Recommendations with Recurrent Neural Networks](https://mlsanity.com/paper/1511.06939/) (2015)
- [RecSim: A Configurable Simulation Platform for Recommender Systems](https://mlsanity.com/paper/1909.04847/) (2019)
- [KuaiRand: An Unbiased Sequential Recommendation Dataset with Randomly Exposed Videos](https://mlsanity.com/paper/2208.08696/) (2022)
- [Alleviating Matthew Effect of Offline Reinforcement Learning in Interactive Recommendation](https://mlsanity.com/paper/2307.04571/) (2023)
- [Future-Conditioned Recommendations with Multi-Objective Controllable Decision Transformer](https://mlsanity.com/paper/2501.07212/) (2025)
- [Goal-Conditioned Supervised Learning for Multi-Objective Recommendation](https://mlsanity.com/paper/2412.08911/) (2024)
- [Offline A/B Testing of Slate Recommendation Systems with LLMs: Reducing the Dependency on Pre-Collected User Interaction Data](https://mlsanity.com/paper/2511.04541/) (2025)

## Abstract (arXiv)

Sequential recommendation optimizes which items to rank, while each displayed slate also shapes subsequent feedback and user state. We study how a trained ranker can support decisions about these future consequences. We introduce UA-TWM, a utility-anchored world-model interface that constructs nearby slate actions, estimates their target-relevant consequences, and selects an alternative subject to utility constraints. The reference slate serves as a fallback when no alternative qualifies. A logged-replay instantiation combines utility and target-gain estimates with calibrated failure-risk prediction; a closed-loop instantiation uses one-step state-action prediction and updates its decisions after observed feedback. We evaluate transfer across twelve sequential backbones on MovieLens-25M and KuaiRand-Pure, and repeated target-directed interaction in KuaiSim. Attaching the interface improves Recall@20, NDCG@20, and future-state alignment for every matched logged backbone. Selection ablations reveal the utility and risk costs of aggressive target pursuit, while closed-loop diagnostics isolate the contribution of action-conditioned prediction. Local consequence modeling thus enables target-aware selection around a trained sequential ranker.