Search papers, labs, and topics across Lattice.
University of California, Riverside, UC Riverside
2
0
5
3
Train smarter, not harder: DSL unlocks 4x faster non-autoregressive generation by teaching masked diffusion models to self-correct more efficiently.
LLMs betray their jailbreaking susceptibility in their hidden activations, allowing for lightweight detection and even real-time disruption of attacks.
