Search papers, labs, and topics across Lattice.
3
0
6
Watermarking can rival traditional membership inference methods, achieving comparable detection performance under specific conditions of subset exposure.
LCLMs redefine the efficiency of long-context inference, achieving superior compression without sacrificing model quality.
Today's best LLMs fail spectacularly at long-horizon reasoning, achieving under 10% accuracy on a new benchmark designed to isolate this critical capability.
