Search papers, labs, and topics across Lattice.
The Chinese University of Hong Kong
5
0
8
Sparsification in attention mechanisms can drastically alter content influence, with higher compression ratios leading to surprising shifts in model output that standard accuracy metrics overlook.
Renormalization can redefine how we understand and address the sim-to-real gap in robotics by leveraging effective parameters that capture omitted dynamics.
A transparent probe-success rule boosts robot policy selection success rates by over 14 percentage points, revealing the hidden power of pre-deployment evaluations.
Agon reveals that machine-driven research can scale effectively while exposing critical failure modes that still require human oversight.
Multi-agent orchestration prompting is critically under-evaluated, with only 14.9% of models passing the new PerspectiveGap benchmark.
