Search papers, labs, and topics across Lattice.
3
0
6
4
MemDreamer narrows the performance gap with human experts in long video understanding to just 3.7 points while processing only 2% of the full context.
Streaming reasoning steps can boost multi-agent system performance by 7.3 percentage points on average, revealing a new dimension for scaling effectiveness and efficiency.
Current MLLMs struggle with fine-grained spatial reasoning, achieving only 37.2 F1 on challenging tasks compared to human performance of 84.0 F1.
