Search papers, labs, and topics across Lattice.
3
0
5
TriAgent achieves an impressive F1 score of ~0.87 while saving $9.3M annually by intelligently routing financial sentiment queries based on contextual granularity.
Despite high agreement in predictions, LLMs underperform against betting markets, revealing stark differences in decision-making quality and self-awareness.
Diffusion LLMs can achieve up to 6.1x higher throughput than autoregressive models by dynamically adjusting decoding granularity based on real-time load, a feat unattainable with fixed-block approaches.
