Search papers, labs, and topics across Lattice.
2
0
5
Training on the entire deployed matrix can unlock up to 81.4% of a model's capacity that was previously unreachable, leading to unprecedented performance gains in low-rank distillation.
Open-source Arabic-DeepSeek-R1 dethrones GPT-5.1 on Arabic language benchmarks, proving that culturally-informed adaptation of sparse MoE models can leapfrog proprietary giants without massive pretraining costs.
