Search papers, labs, and topics across Lattice.
6
0
8
4
MSCENet achieves unprecedented accuracy in detecting anomalies in multivariate time series by integrating advanced spatio-temporal learning with graph-based correlation modeling.
TREK transforms the way models tackle challenging prompts by expanding their exploration support, leading to substantial performance gains even in the hardest task scenarios.
Stop wasting compute on easy and impossible examples: PACED distillation focuses your student model's training on the sweet spot where it actually learns.
Quadruped robots can now continuously jump across rough, lunar-like terrain using only onboard sensors, thanks to a new dual-horizon control model and a clever hardware-in-the-loop validation platform.
Reasoning models aren't just verbose, they're actively *harmed* by their own verbosity, but a simple self-distillation trick can compress their outputs by up to 59% while boosting accuracy by up to 16 points.
Overconfident errors in RLVR monopolize probability mass and suppress exploration, but a confidence-aware penalty fixes this and boosts mathematical reasoning performance.
