Papers/2609.01609
🧪 Test?View on arXiv

DiDrive: A Risk-Aware Hierarchical Diffusion Framework for Safe Offline Reinforcement Learning in Autonomous Driving

Author1, Author2, Author3, Author4, Author5

offline reinforcement learningautonomous drivingrisk-awarediffusion models
2609.01609
Builder Relevance
80%
2h ago

Abstract

DiDrive proposes a distribution-guided offline diffusion framework to enhance safety in autonomous driving by addressing challenges in offline reinforcement learning.

Reality Card

Core Claim

DiDrive achieves an 85% success rate and a 4295.68 average reward in complex traffic scenarios, outperforming existing baselines.

Method / Result

85% success rate in high-density traffic scenarios with 60 vehicles.

Limitations

The complexity of the framework may pose challenges for reproducibility in different environments.

Paper to code

Verified implementation resources so builders can test the paper’s claims instead of stopping at the abstract.

No verified implementation link has been attached yet. AIBuzzHub will keep this panel separate from unverified search results.
← Back to all papers