Search papers, labs, and topics across Lattice.
7
0
10
0
Watermarking LLM-agent trajectories just got a major upgrade鈥擳RACE achieves near-perfect detection without sacrificing performance, even under aggressive adversarial conditions.
Achieving nearly three times faster inference in speech synthesis without sacrificing speaker similarity could redefine efficiency benchmarks in the field.
Re-ranking can make or break user engagement, and GR2 boosts performance by over 18% by harnessing the power of LLMs in ways previously unexplored.
Wastewater-first monitoring can now optimize decision-making under uncertainty, significantly improving cost-performance while ensuring scientific rigor.
Training LLMs to verify their own math proofs is now twice as fast, thanks to a clever fusion of generation and verification into a single training pass.
LLMs can learn musicality without human annotation by aligning them to automatically generated preference datasets derived from rule-based musical constraints.
Despite users preferring human-created videos, AI-generated content can achieve similar overall engagement on video platforms by flooding the system with sheer volume.
