Search papers, labs, and topics across Lattice.
Zhejiang University
2
0
4
4
Forget OCR: DocRetriever achieves state-of-the-art multimodal document retrieval by cleverly combining layout-aware sparse embeddings with a reasoning-augmented reranker.
LALMs struggle to keep track of sparse events across hours of audio, unlike humans who excel at this, revealing a key memory persistence bottleneck.
