SimpleMemVLA

VLA from HUST, Tsinghua, ModelBest, Renmin University and BAAI that uses the backbone’s native timestamped video history as memory. The parameter count is the Hugging Face safetensors total of the RoboDojo checkpoint.

SimpleMemVLA Team · Sep 2, 2026 · VLAReport an error
9.3%Dojo Sim · #37 of 71
6.3BParameters
1Benchmarks reported
Sep 2, 2026Released

Benchmark results1

#37 of 719.3%
Not reported on 14 benchmarks
Wrong or missing number? Report an error