Search papers, labs, and topics across Lattice.
Beijing Institute of Technology, Zhuhai ;
5
0
8
NA-LoRA reveals that adapting low-rank updates with a focus on gate channel responsiveness can significantly enhance model fine-tuning performance.
EO-WM achieves unprecedented accuracy in forecasting vegetation response to extreme weather, outperforming traditional models by effectively modeling uncertainty and cumulative stress.
Bridging the gap in multi-camera depth prediction, SurroundNEXO achieves a 33.2% reduction in single-view error by rethinking how we leverage ego-centric geometry.
SPD achieves unprecedented decoding speed by processing multiple tokens in parallel while eliminating latency bubbles, setting a new standard for LLM inference efficiency.
UAV-VLN is still far from real-world deployment due to challenges like the sim-to-real gap and linguistic ambiguity, highlighting opportunities for research in areas like multi-agent coordination.
