Search papers, labs, and topics across Lattice.
8
0
10
10
Achieving precise timing control in audio-visual generation without model retraining could revolutionize how we synchronize speech and visuals in AI applications.
A single successful attack can reveal persistent vulnerabilities in LLM safeguards, challenging the assumption that improved local scores equate to enhanced safety in deployment.
PURA achieves over three times the message match rate of existing unbiased watermarking methods, all while preserving text quality and speed.
A safety trilemma reveals that LLMs cannot balance utility, safety, and access when evidence is copyable, exposing a critical vulnerability in current safeguards.
Relying only on affinity can lead to failures in noisy environments, but ANFI's dual approach to neighbor interactions significantly boosts robustness in person re-ID tasks.
VisCo not only compresses visual tokens more effectively than previous methods but also enhances model performance through innovative use of memory tokens.
Stop letting semantics dictate composition: Composer unlocks semantic-agnostic control over image aesthetics, letting you transfer and plan compositions with unprecedented precision.
LLM agent distillation leads to surprisingly high rates of behavioral mimicry, with some student models exhibiting tool-use habits *more* similar to their teachers than the teacher's own family members.
