Search papers, labs, and topics across Lattice.
Xiaomi EV
4
0
7
2
Dense neural networks are choking on sparse recommendation data, but SSR's explicit sparsity unlocks continuous performance gains where dense models saturate.
Naturalness-based data selection, a common technique for curating LLM reasoning datasets, systematically favors longer, lower-quality reasoning chains due to a previously unnoticed "step length confounding" effect.
A Transformer-based ranking model can boost e-commerce orders by 6.35% while halving latency, thanks to optimizations targeting feature sparsity and low label density.
Forget interaction-driven next-item prediction: SIGMA uses instruction-following and semantic grounding to create a generative recommender that adapts to evolving trends and diverse tasks on AliExpress.
