What Probing Reveals about Autonomous Driving: Linking Internal Prediction Errors to Ego Planning

• URL: http://arxiv.org/abs/2606.31106v1
• Date: Tue, 30 Jun 2026 04:04:01 GMT
• Title: What Probing Reveals about Autonomous Driving: Linking Internal Prediction Errors to Ego Planning
• Abstract summary: Large-scale datasets and simulators have enabled improvements in driving policies that appear safe and robust.<n>But strong performance in nominal scenarios can still mask flawed reasoning and unsafes.<n>We focus on probing for forms of prediction, where surrounding vehicles will move next, and planning, understanding how to generate safe trajectories.
• Score: 4.0952543282008795
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large-scale datasets and fast simulators have enabled improvements in driving policies that appear safe and robust, yet strong performance in nominal scenarios can still mask flawed reasoning and unsafe heuristics. Summary scores from closed-loop simulators do not give significant insight into the policy, making it difficult to determine whether they truly predict the motion of surrounding vehicles, how the ego vehicle generates future plans, or whether they merely rely on brittle heuristics that happen to succeed in nominal scenarios. To better understand the limits and weaknesses of driving policies, we focus on probing for forms of prediction, i.e., where surrounding vehicles will move next, and planning, i.e., understanding how to generate safe trajectories. We focus on these two capabilities because they reflect behaviors expected of effective driving policies, and use their presence or absence to assess policy quality across data-driven behavior cloning and simulation-driven reinforcement learning policies. To evaluate the presence of these capabilities, we investigate them as a function of scale, asking whether the closed-loop gains from larger datasets and longer simulation training reflect stronger prediction and planning or merely better behavioral heuristics. We use linear probing and targeted perturbations in both imitation learning and reinforcement learning models to track when these internal signals emerge, plateau, or fail. Despite good closed-loop performance, policies often fail to form timely surrounding-vehicle predictions during near-collision events, revealing a limitation in the predictive signals available for ego planning. Finally, causal intervention shows that correcting mistaken predictions improves ego planning toward safer trajectories.

Related papers

• From Attacks to Curricula: Learnability-Guided Adversarial Training for Safe Autonomous Driving [56.30087557121323]
  AlignADV is a learnability-guided closed-loop adversarial training framework.<n>It converts adversarial scenarios into resolvable and capability-aligned curricula.<n> Experiments show it reduces training steps by up to 40.6 percent.
  arXiv  Detail & Related papers  (2026-06-12T02:13:55Z)
• Planning with Adaptive World Models for Autonomous Driving [50.4439896514353]
  We present nuPlan, a real-world motion planning benchmark that captures multi-agent interactions.<n>We learn to model such unique behaviors with BehaviorNet, a graph convolutional neural network (GCNN)<n>We also present AdaptiveDriver, a model-predictive control (MPC) based planner that unrolls different world models conditioned on BehaviorNet's predictions.
  arXiv  Detail & Related papers  (2024-06-15T18:53:45Z)
• Uncertainty-Aware DRL for Autonomous Vehicle Crowd Navigation in Shared Space [3.487370856323828]
  This work introduces an integrated prediction and planning approach that incorporates the uncertainties of predicted pedestrian states in the training of a model-free DRL algorithm. A novel reward function encourages the AV to respect pedestrians' personal space, decrease speed during close approaches, and minimize the collision probability with their predicted paths. Results show a 40% decrease in collision rate and a 15% increase in minimum distance to pedestrians compared to the state of the art model that does not account for prediction uncertainty.
  arXiv  Detail & Related papers  (2024-05-22T20:09:21Z)
• SAFE-SIM: Safety-Critical Closed-Loop Traffic Simulation with Diffusion-Controllable Adversaries [94.84458417662407]
  We introduce SAFE-SIM, a controllable closed-loop safety-critical simulation framework. Our approach yields two distinct advantages: 1) generating realistic long-tail safety-critical scenarios that closely reflect real-world conditions, and 2) providing controllable adversarial behavior for more comprehensive and interactive evaluations. We validate our framework empirically using the nuScenes and nuPlan datasets across multiple planners, demonstrating improvements in both realism and controllability.
  arXiv  Detail & Related papers  (2023-12-31T04:14:43Z)
• The Integration of Prediction and Planning in Deep Learning Automated Driving Systems: A Review [43.30610493968783]
  We review state-of-the-art deep learning-based planning systems, and focus on how they integrate prediction. We discuss the implications, strengths, and limitations of different integration principles.
  arXiv  Detail & Related papers  (2023-08-10T17:53:03Z)
• What Truly Matters in Trajectory Prediction for Autonomous Driving? [25.700880872677256]
  Trajectory prediction plays a vital role in the performance of autonomous driving systems. A significant disparity exists between the accuracy of predictors on fixed datasets and driving performance when the predictors are used downstream for vehicle control.
  arXiv  Detail & Related papers  (2023-06-27T01:29:20Z)
• Self-Aware Trajectory Prediction for Safe Autonomous Driving [9.868681330733764]
  Trajectory prediction is one of the key components of the autonomous driving software stack. In this paper, a self-aware trajectory prediction method is proposed. The proposed method performed well in terms of self-awareness, memory footprint, and real-time performance.
  arXiv  Detail & Related papers  (2023-05-16T03:53:23Z)
• P4P: Conflict-Aware Motion Prediction for Planning in Autonomous Driving [28.948224519638913]
  We evaluate state-of-the-art predictors through novel conflict-related metrics. We propose a simple but effective alternative that combines a physics-based trajectory generator and a learning-based predictor. Our predictor, P4P, achieves superior performance over existing learning-based predictors in realistic interactive driving scenarios.
  arXiv  Detail & Related papers  (2022-11-03T07:51:40Z)
• AdvDO: Realistic Adversarial Attacks for Trajectory Prediction [87.96767885419423]
  Trajectory prediction is essential for autonomous vehicles to plan correct and safe driving behaviors. We devise an optimization-based adversarial attack framework to generate realistic adversarial trajectories. Our attack can lead an AV to drive off road or collide into other vehicles in simulation.
  arXiv  Detail & Related papers  (2022-09-19T03:34:59Z)
• Safety-aware Motion Prediction with Unseen Vehicles for Autonomous Driving [104.32241082170044]
  We study a new task, safety-aware motion prediction with unseen vehicles for autonomous driving. Unlike the existing trajectory prediction task for seen vehicles, we aim at predicting an occupancy map. Our approach is the first one that can predict the existence of unseen vehicles in most cases.
  arXiv  Detail & Related papers  (2021-09-03T13:33:33Z)
• Contingencies from Observations: Tractable Contingency Planning with Learned Behavior Models [82.34305824719101]
  Humans have a remarkable ability to make decisions by accurately reasoning about future events. We develop a general-purpose contingency planner that is learned end-to-end using high-dimensional scene observations. We show how this model can tractably learn contingencies from behavioral observations.
  arXiv  Detail & Related papers  (2021-04-21T14:30:20Z)
• The Importance of Prior Knowledge in Precise Multimodal Prediction [71.74884391209955]
  Roads have well defined geometries, topologies, and traffic rules. In this paper we propose to incorporate structured priors as a loss function. We demonstrate the effectiveness of our approach on real-world self-driving datasets.
  arXiv  Detail & Related papers  (2020-06-04T03:56:11Z)
• Perception as prediction using general value functions in autonomous driving applications [5.071770433010771]
  Perception as prediction learns data-driven predictions relating to the impact of actions on the agent's perception of the world. We demonstrate perception as prediction by learning to predict an agent's front safety and rear safety with GVFs. We show that these predictions can be used to produce similar control behavior to an LQR-based controller in an adaptive cruise control problem.
  arXiv  Detail & Related papers  (2020-01-24T17:33:06Z)

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.