Search papers, labs, and topics across Lattice.
4
0
5
DSG achieves 91% lower search costs while maintaining competitive accuracy, revolutionizing how LLM agents handle real-time search and reasoning.
Forget optimizing solely for clicks: injecting LLM-derived semantic relevance into e-commerce ranking models can boost search quality without sacrificing engagement.
Stop guessing how to improve your multi-agent shopping assistant: this paper offers a practical blueprint, evaluation rubric, and prompt optimization strategies (Sub-agent GEPA and MAMuT GEPA) for production-scale systems.
Achieve statistically significant gains in e-commerce search relevance and business impact by explicitly modeling graded relevance levels during contrastive embedding training.
