Bench-2Drive

The 'Bench-2Drive' dataset/benchmark contains diverse driving scenarios and is used to evaluate the performance of autonomous driving systems in generating and refining trajectory proposals through iterative decision-making.

Papers using Bench-2Drive (3)

CRAFT: Counterfactual-to-Interactive Reinforcement Fine-Tuning for Driving Policies2026EvaDrive: Evolutionary Adversarial Policy Optimization for End-to-End Autonomous Driving2025Raw2Drive: Reinforcement Learning with Aligned World Models for End-to-End Autonomous Driving (in CARLA v2)2025