Search papers, labs, and topics across Lattice.
Shanghai Jiao Tong University
1
0
3
Current vision-language models struggle with process understanding in robotic manipulation, but targeted post-training can yield significant improvements.
