Search papers, labs, and topics across Lattice.
4
0
8
1
The results suggest that dense to MoE adaptation with dynamic expert deactivation is a practical direction for reducing active VLA model size without severe performance loss.
Generating agentic data isn't just about quantity; it's about crafting valid, informative experiences that evolve alongside agents and their environments.
Synthesized egocentric videos can elevate real-robot task success rates by over 10% through enhanced training data.
Achieving over 2x speedup in VLA inference without sacrificing performance could revolutionize real-time robotic applications.
