Search papers, labs, and topics across Lattice.
3
0
6
1
LLMs can now generate mobile kernels that outperform native libraries 27.4% of the time, thanks to a new multi-agent system that tackles compilation failures and hallucinations.
Finally, a neural audio codec that can stream high-fidelity music at standard bitrates (32-128kbps) with superior quality.
Diffusion LMs can now beat autoregressive models at CUDA kernel generation, thanks to a new dataset and reinforcement learning approach.
