Search papers, labs, and topics across Lattice.
3
0
7
8
Token-level memorization asymmetry in diffusion models reveals a critical vulnerability, enabling effective membership inference and PII extraction.
LLM agents struggle to maintain performance in multi-day collaborative tasks, dropping significantly after just one environmental update, revealing a critical gap in adaptation to evolving real-world conditions.
Now you can audit black-box LLM APIs for cheating (model substitution, overbilling) with <1% overhead, using verifiable computation.
