Frontend
ReactBench-AD: Do Driving VLAs and World-Action Models React to Imminent Hazards in Their Path?
arXiv cs.RO0 views
arXiv:2611.20011v1 Announce Type: cross Abstract: Vision-Language-Action models (VLAs) and World-Action Models (WAMs) have emerged in autonomous driving with the promise of strong generalization to long-tail scenarios. Although these models achieve state-of-the-art results on autonomous driving benchmarks, most existing benchmarks are dominated by nominal scenarios and therefore cannot test this promise. To test one such long-tail capability that is critical for safety, we introduce ReactBench-AD, a benchmark that measures how driving models react to objects that appear in the ego vehicle's path. Built on real NAVSIM and PhysicalAI-Autonomous-Vehicles scenes, ReactBench-AD composites a crossing pedestrian, a standing pedestrian, or a stationary car consistently across all camera views and history frames. Each edited scene thus forms an exact, pixel-level counterfactual pair with its original, which allows us to isolate the causal effect of the inserted object on the predicted trajectory. Evaluating four VLAs and two WAMs, we find that they react weakly, often barely at all. Finally, we test how easily one of the evaluated VLAs, ReCogDrive, can be taught to brake hard for imminent hazards: a short fine-tuning stage on edited scenes substantially improves its reactions, with the learned behavior generalizing to unseen objects and fully real scenes, at minimal cost to nominal driving. These results call into question the out-of-the-box long-tail capabilities of these models, but suggest that they can be a strong foundation for acquiring them with little additional training, highlighting the value of real or synthetic data that demonstrates the long-tail behaviors nominal datasets rarely contain. Code and data will be made publicly available.
Source:https://arxiv.org/abs/2611.20011PreviousThe Flip Test: Video World Models Fail to Learn the World StateNextD'BEAST: Logged Bad Experiences as Explanation for Autonomous Driving Agents
Related posts
How We Built 1,200 SEO Landing Pages for a Local Plumber in 4 Hours (Programmatic SEO + Supabase)
0dev.to
55 of 55 URLs Returned 522. The Site Was Fine — My Probe Was Fetching Its Own Zone
0dev.to
Negative marking in Google Forms with Apps Script
0dev.to
Twelve Attempts to Reply. Nothing Sent.
0dev.to
Why We Replaced Our Freelance WordPress Stack With Lovable + Supabase (6-Month Numbers)
0dev.to
Source Deduplication Patterns for a Healthcare Appointment Retrieval Architecture
0dev.to