Learning Safe Agent Behaviour from Human Preferences and Justifications via World Models

• URL: http://arxiv.org/abs/2607.13172v1
• Date: Tue, 14 Jul 2026 18:22:14 GMT
• Title: Learning Safe Agent Behaviour from Human Preferences and Justifications via World Models
• Abstract summary: We introduce DROPJ, a human-centred method for both safe training and deployment.<n>We first learn a world model (a learned simulator) from a dataset of prior real-world trajectories.<n>A human then plays the game in this learned simulator to extract several informative simulated trajectories.<n>We then train a reward model from these justified preferences and use it, together with the world model, to directly deploy the agent.
• Score: 10.47361214312137
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: We address the problem of safely training an agent policy and deploying a good and safe policy, in settings where the environment dynamics are unknown and no suitable reward function is available. In the context of safety-critical environments, we consider traditional reinforcement learning impractical and resort to the resource of human input. We introduce DROPJ, a human-centred method for both safe training and deployment. We first learn a world model (a learned simulator) from a dataset of prior real-world trajectories. A human then plays the game in this learned simulator to extract several informative simulated trajectories. From these, we sample pairs of simulated trajectory segments and elicit from a human their preference over these segments, as well as a reason (justification) for their choice. We then train a reward model from these justified preferences and use it, together with the world model, to directly deploy the agent using model predictive control. Running real-user experiments, we find that generating informative simulated trajectories from a user significantly reduces the computational cost during training compared to other strategies, and can also improve the performance during deployment. In the context of training within a learned simulator, we show that the use of preferences rather than other types of feedback substantially improves the performance during deployment. We further demonstrate that safety justifications accompanying preferences can significantly enhance safety or prioritise user-prescribed aspects of safety associated with them during deployment.

Related papers

• Advantage-Driven Explicit Memory for Social Navigation [12.557252183063]
  We propose a new navigation agent equipped with non-parametric memory which explicitly indexes prior steps leading to critical events.<n>In the context of social navigation, we show that this improves the agent's capability to retain sparse, high-cost failures.<n>We train our agent in simulation with a combination of rendering and non-visual crowd simulation and show that the agent is robust with respect to OOD social behavior.
  arXiv  Detail & Related papers  (2026-08-26T10:27:08Z)
• Autonomous Continual Learning of Computer-Use Agents for Environment Adaptation [57.65688895630163]
  We introduce ACuRL, an Autonomous Curriculum Reinforcement Learning framework that continually adapts agents to specific environments with zero human data.<n>Our method effectively enables both intra-environment and cross-environment continual learning, yielding 4-22% performance gains without forgetting existing environments.
  arXiv  Detail & Related papers  (2026-02-10T23:06:02Z)
• Semantic-Metric Bayesian Risk Fields: Learning Robot Safety from Human Videos with a VLM Prior [10.193047117743221]
  Humans interpret safety not as a binary signal but as a continuous, context- and spatially-dependent notion of risk.<n>This work proposes a framework for extracting implicit human risk models supervised directly from safe human demonstration videos and VLM common sense.
  arXiv  Detail & Related papers  (2025-12-09T04:19:58Z)
• SPACeR: Self-Play Anchoring with Centralized Reference Models [50.55045557371374]
  Sim agent policies are realistic, human-like, fast, and scalable in multi-agent settings.<n>Recent progress in imitation learning with large diffusion-based or tokenized models has shown that behaviors can be captured directly from human driving data.<n>We propose SPACeR, a framework that leverages a pretrained tokenized autoregressive motion model as a central reference policy.
  arXiv  Detail & Related papers  (2025-10-20T19:53:02Z)
• Autonomous Vehicle Controllers From End-to-End Differentiable Simulation [57.278726604424556]
  We propose a differentiable simulator and design an analytic policy gradients (APG) approach to training AV controllers.<n>Our proposed framework brings the differentiable simulator into an end-to-end training loop, where gradients of environment dynamics serve as a useful prior to help the agent learn a more grounded policy.<n>We find significant improvements in performance and robustness to noise in the dynamics, as well as overall more intuitive human-like handling.
  arXiv  Detail & Related papers  (2024-09-12T11:50:06Z)
• ReGentS: Real-World Safety-Critical Driving Scenario Generation Made Stable [88.08120417169971]
  Machine learning based autonomous driving systems often face challenges with safety-critical scenarios that are rare in real-world data. This work explores generating safety-critical driving scenarios by modifying complex real-world regular scenarios through trajectory optimization. Our approach addresses unrealistic diverging trajectories and unavoidable collision scenarios that are not useful for training robust planner.
  arXiv  Detail & Related papers  (2024-09-12T08:26:33Z)
• Real-World Humanoid Locomotion with Reinforcement Learning [92.85934954371099]
  We present a fully learning-based approach for real-world humanoid locomotion. Our controller can walk over various outdoor terrains, is robust to external disturbances, and can adapt in context.
  arXiv  Detail & Related papers  (2023-03-06T18:59:09Z)
• Off Environment Evaluation Using Convex Risk Minimization [0.0]
  We propose a convex risk minimization algorithm to estimate the model mismatch between the simulator and the target domain. We show that this estimator can be used along with the simulator to evaluate performance of an RL agents in the target domain.
  arXiv  Detail & Related papers  (2021-12-21T21:31:54Z)
• Reactive and Safe Road User Simulations using Neural Barrier Certificates [9.961324632236499]
  We propose a reactive agent model which can ensure safety without comprising the original purposes. Our learned road user simulation models can achieve a significant improvement in safety. Our learned reactive agents are shown to generalize better to unseen traffic conditions.
  arXiv  Detail & Related papers  (2021-09-14T13:45:37Z)
• TrafficSim: Learning to Simulate Realistic Multi-Agent Behaviors [74.67698916175614]
  We propose TrafficSim, a multi-agent behavior model for realistic traffic simulation. In particular, we leverage an implicit latent variable model to parameterize a joint actor policy. We show TrafficSim generates significantly more realistic and diverse traffic scenarios as compared to a diverse set of baselines.
  arXiv  Detail & Related papers  (2021-01-17T00:29:30Z)
• Guided Uncertainty-Aware Policy Optimization: Combining Learning and Model-Based Strategies for Sample-Efficient Policy Learning [75.56839075060819]
  Traditional robotic approaches rely on an accurate model of the environment, a detailed description of how to perform the task, and a robust perception system to keep track of the current state. reinforcement learning approaches can operate directly from raw sensory inputs with only a reward signal to describe the task, but are extremely sample-inefficient and brittle. In this work, we combine the strengths of model-based methods with the flexibility of learning-based methods to obtain a general method that is able to overcome inaccuracies in the robotics perception/actuation pipeline.
  arXiv  Detail & Related papers  (2020-05-21T19:47:05Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.