Search papers, labs, and topics across Lattice.
Beijing University of Posts and Telecommunications
3
0
6
Adaptive weighting in model merging can drastically improve multilingual reasoning performance, outperforming traditional methods across 21 languages.
Low-light 3D scene reconstruction from sparse views is now possible with feed-forward Gaussian splatting, thanks to a new method that directly predicts clean 3D Gaussians even when context views are noisy.
Unleashing an LLM's inner creativity or laser-sharp logic is now as simple as turning a knob, thanks to a new distribution-matching method that avoids heuristic rewards.
