Search papers, labs, and topics across Lattice.
3
0
6
8
Pixel Linguist II not only sets new benchmarks in multilingual visual text understanding but also maintains performance under 80% visual token compression.
Contrastive Policy Optimization reveals that distinguishing between correct and incorrect responses can significantly enhance reinforcement learning performance.
LLMs can learn to tailor their queries for different retrievers, revealing that one-size-fits-all strategies are ineffective in retrieval-augmented generation.
