Search papers, labs, and topics across Lattice.
and AI Institute, Equal Advising
4
0
6
Achieving a 35% success rate in real-world robotic tasks using only synthetic data marks a groundbreaking advance in sim-to-real transfer for world-action models.
PRISM achieves up to 15% performance gains in visuomotor tasks by leveraging short-term memory, challenging the notion that immediate sensory input suffices for complex decision-making.
Forget brittle imitation learning: Q2RL unlocks robust on-robot reinforcement learning by distilling a Q-function from Behavior Cloning and intelligently gating between imitation and RL based on Q-value estimates.
Robots can now learn complex manipulation tasks from scratch using only video and language, bypassing the need for hand-engineered reward functions, demonstrations, or even task-specific tuning.
