Search papers, labs, and topics across Lattice.
Affiliation:
6
0
6
2
OrthoSkillVLA preserves prior skills in pretrained VLA models while seamlessly integrating new ones, outperforming traditional methods in both simulated and real-world scenarios.
GRACE accelerates real-time ad retrieval by improving eligibility matching rates and drastically reducing latency, making it feasible to generate thousands of ads on demand.
Re-ranking can make or break user engagement, and GR2 boosts performance by over 18% by harnessing the power of LLMs in ways previously unexplored.
Achieving state-of-the-art occupancy prediction while using only 2D images, Occ-VLM bridges the gap between 2D semantics and 3D understanding without the need for complex 3D inputs.
Current multimodal LLMs struggle with UI-based reasoning, but the new UI-UX model achieves a remarkable 0.7963 accuracy on the UXBench benchmark, setting a new standard.
Robot control gets a whole lot faster: ProbeFlow slashes action decoding latency by 14.8x in Vision-Language-Action models, all without retraining.
