Search papers, labs, and topics across Lattice.
Zhejiang University
1
0
3
Dynamo achieves a remarkable 5.6% average accuracy boost in visual reasoning tasks without retraining, revolutionizing how VLMs adapt in real-time.
