Search papers, labs, and topics across Lattice.
4
5
6
23
Separating semantic and linguistic features can drastically enhance multilingual retrieval performance, especially in low-resource settings.
CroSearch-R1 reveals that integrating cross-lingual knowledge through a dynamic retrieval strategy can substantially enhance the performance of Retrieval-Augmented Generation systems.
KDFlow unlocks 6x faster LLM knowledge distillation by decoupling student training and teacher inference, using hidden state transfer to minimize communication overhead.
Token-level alignment, powered by a novel distillation approach, lets LLMs learn faster and better by avoiding the pitfalls of response-level reward optimization.
