Search papers, labs, and topics across Lattice.
6
0
11
16
Achieving a 57.6% success rate on RoboCasa365, Xiaomi-Robotics-1 sets a new standard for vision-language-action models in real-world robotic manipulation.
A novel RL framework for Flow-Matching TTS achieves faster convergence and superior audio quality, outperforming traditional methods.
JourneyFormer transforms Airbnb's search ranking by effectively modeling complex guest journeys, leading to substantial gains in business performance.
GF-DiT achieves up to 6.01× throughput improvement and 95% latency reduction by dynamically adapting GPU parallelism in response to workload demands.
Forget heuristics – this work gives provable conditions for *when* and *how* auxiliary data actually improve generalization in transfer learning.
By aligning latent representations with multiple visual foundation models, FRAPPE offers a more scalable and data-efficient way to imbue generalist robotic policies with robust world-awareness.
