Search papers, labs, and topics across Lattice.
3
4
6
4
Reusing an MLLM's causal context as robot memory boosts performance on complex tasks, achieving a remarkable 75.54% accuracy on RoboMME.
Blindly adding images to fact-checking hurts more than it helps; this adaptive framework knows when to look.
Pokemon, not just a childhood game, emerges as a surprisingly effective benchmark for AI, revealing critical gaps in LLMs and RL agents that existing benchmarks miss.
