1. [Publications](/publications)
2. Beyond Behavior Cloning in Autonomous Driving: a Survey of Closed-Loop Training Techniques
 
 # Beyond Behavior Cloning in Autonomous Driving: a Survey of Closed-Loop Training Techniques

  ![](/sites/default/files/styles/wide/public/publications/cl_training_v3.png?itok=6yozI_In)

 Behavior cloning, the dominant approach for training autonomous vehicle (AV) policies, suffers from a fundamental gap: policies trained open-loop on temporally independent samples must operate in closed-loop where actions influence future observations. This mismatch can cause covariate shift, compounding errors, and poor interactive behavior, among other issues. Closed-loop training mitigates the problem by exposing policies to the consequences of their actions during training. However, the recent shift to end-to-end ("sensor to action'') systems has made closed-loop training significantly more complex, requiring costly high-dimensional rendering and managing sim-to-real gaps. This survey presents a comprehensive taxonomy of closed-loop training techniques for end-to-end driving, organized along three axes: action generation (policy rollouts vs. perturbed demonstrations); environment response generation (real-world data collection, AV simulation, generative video and latent world models); and training objectives (closed-loop imitation, reinforcement learning, and their combinations). We analyze key trade-offs along each axis: on-policy vs. on-expert action generation, environment fidelity vs. cost, and expert vs. reward-based training objectives; as well as coupling factors, such as rollout deviation from the policy, expert, and real world logs; and data type, throughput, and latency requirements. The analysis reveals gaps between current research and industry practice, and points to promising directions for future work.



 ## Authors



[Peter Karkus](/person/peter-karkus)

[Maximilian Igl](/person/maximilian-igl)

[Yuxiao Chen](/person/yuxiao-chen)

Kashyap Chitta (NVIDIA)

Jef Packer (NVIDIA)

[Bertrand Douillard](/person/bertrand-douillard)

[Thomas Tian](/person/thomas-tian)

Alexander Naumann (NVIDIA)

Guillermo Garcia-Cobo (NVIDIA)

Shuhan Tan (NVIDIA)

[Alperen Degirmenci](/person/alperen-degirmenci)

Alexander Popov (NVIDIA)

Nikolai Smolyanskiy (NVIDIA)

Urs Muller (NVIDIA)

[Boris Ivanovic](/person/boris-ivanovic)

[Marco Pavone](/person/marco-pavone)

 

 

 ## Publication Date



Friday, December 5, 2025

 

 ## Research Area



[Artificial Intelligence and Machine Learning ](/research-area/machine-learning-artificial-intelligence)

[Autonomous Vehicles](/research-area/autonomous-vehicles)

[Computer Vision](/research-area/computer-vision)

[Generative AI](/research-area/generative-ai)

[Physical AI](/research-area/physical-ai)

[World Simulation](/research-area/world-simulation)

 

 

 ## Uploaded Files



[beyond\_bc\_survey\_preprint.pdf](https://d1qx31qr3h6wln.cloudfront.net/publications/beyond_bc_survey_preprint.pdf "Open file in new window")6.59 MB