Search papers, labs, and topics across Lattice.
5
0
9
3
A vision-only driving model, distilled from a massive VLA teacher, not only matches but *exceeds* its teacher's performance, proving that there's still headroom in vision-centric architectures for autonomous driving.
Radar point clouds, when enriched with spectral information, can outperform traditional dense range-Doppler spectra, suggesting a path toward more robust and generalizable radar perception models.
LLMs may ace synthetic benchmarks, but they fumble the efficiency test in real-world cloud service scenarios, revealing a critical gap in their readiness for customer-facing applications.
Achieve SOTA LLM alignment in complex technical domains with a fraction of the compute by distilling knowledge into smaller models using a hybrid reward mechanism and targeted data augmentation.
This new OCR model beats Gemini-3.1-Pro and Qwen3-VL-235B on key information extraction, thanks to its clever "Layout-as-Thought" process that recovers layout grounding in end-to-end OCR.
