SafeLIBERO is a benchmark designed to evaluate robotic model performance in complex, safety-critical environments. It extends each LIBERO suite by selecting four representative tasks, with each task ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results