CLEAR: Closed-Loop Reinforcement Learning at Scale for End-to-End Autonomous Driving
Open paper
Benchmark
Model
Rank
Results
bench2drive-on-bench2drive
CLEAR (InternVL2-1B)
–
Driving Score: 85.9
Success Rate: 68.2
bench2drive-on-bench2drive
CLEAR (InternVL3-1B)
–
Driving Score: 86.8
Success Rate: 69.5
Rank counts only results with a code link.