Search papers, labs, and topics across Lattice.
National Taiwan University, Taiwan
1
0
3
9
GPT-4o can't handle commonsense planning with changing object affordances as well as a LoRA-finetuned vision-language model trained for the task.