# Hindsight Preference Replay Improves Preference-Conditioned Multi-Objective Reinforcement Learning

arXiv: https://arxiv.org/abs/2601.11604 · HTML: https://mlsanity.com/paper/2601.11604/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Deep RL algorithms and representations, one of 15 directions in Deep reinforcement learning methods.
- Builds on 7 of its 18 references that are in the map.
- First posted on arXiv 8 Jan 2026.

## Placed in

- Deep reinforcement learning methods / [Deep RL algorithms and representations](https://mlsanity.com/deeprl/deep-rl-algorithms-and-representations/)

## Builds on (most cited first, 7 of 7)

- [Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor](https://mlsanity.com/paper/1801.01290/) (2018)
- [Hindsight Experience Replay](https://mlsanity.com/paper/1707.01495/) (2017)
- [A Practical Guide to Multi-Objective Reinforcement Learning and Planning](https://mlsanity.com/paper/2103.09568/) (2021)
- [A Survey of Multi-Objective Sequential Decision-Making](https://mlsanity.com/paper/1402.0590/) (2014)
- [Preference-Driven Multi-Objective Combinatorial Optimization with Conditional Computation](https://mlsanity.com/paper/2506.08898/) (2025)
- [Optimizing Deep Reinforcement Learning for Adaptive Robotic Arm Control](https://mlsanity.com/paper/2407.02503/) (2024)
- [Advances in Experience Replay](https://mlsanity.com/paper/1805.05536/) (2018)

## Abstract (arXiv)

Multi-objective reinforcement learning (MORL) enables agents to optimize vector-valued rewards while respecting user preferences. CAPQL, a preference-conditioned actor-critic method, achieves this by conditioning on weight vectors w and restricts data usage to the specific preferences under which it was collected, leaving off-policy data from other preferences unused. We introduce Hindsight Preference Replay (HPR), a simple and general replay augmentation strategy that retroactively relabels stored transitions with alternative preferences. This densifies supervision across the preference simplex without altering the CAPQL architecture or loss functions. Evaluated on six MO-Gymnasium locomotion tasks at a fixed 300000-step budget using expected utility (EUM), hypervolume (HV), and sparsity, HPR-CAPQL improves HV in five of six environments and EUM in four of six. On mo-humanoid-v5, for instance, EUM rises from 323!±!125 to 1613!±!464 and HV from 0.52M to 9.63M, with strong statistical support. mo-halfcheetah-v5 remains a challenging exception where CAPQL attains higher HV at comparable EUM. We report final summaries and Pareto-front visualizations across all tasks.