Search papers, labs, and topics across Lattice.
University of California
4
0
8
Entity embeddings outperform traditional encoding methods in high-cardinality fraud detection, achieving a record AUC-ROC score that could redefine best practices in the field.
Depth accuracy varies dramatically across cameras, with biases ranging from 50 mm to over 1400 mm, highlighting the critical need for informed camera selection in healthcare monitoring.
Agentic RL can now beat proprietary LLMs and torch.compile in the challenging domain of CUDA kernel generation, achieving up to 40% speedups on hard tasks.
Diffusion models can achieve a 10x performance boost in real-world autonomous driving when optimized for trajectory representation, loss space, and augmented with reinforcement learning.
