Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
Affiliation:
6
0
4
0
A standardized protocol for LLM brand recommendation auditing reveals that just 15 iterations can significantly enhance reliability, challenging existing practices in the field.
PersonaGen-1M reveals that 78.3% of buyer queries are informational, providing a crucial lens into how generative engines prioritize brand recommendations.
Query language alone can determine whether local suppliers are included in product recommendations, regardless of the user's location.
Citation type, not volume, is the key predictor of whether language models name individuals, revealing a nuanced understanding of AI-driven visibility.
Recommendation concentration is moderate, with only 41.6% agreement on top brands across different models, defying the winner-takes-all assumption in AI recommendations.
AI-generated brand reputations are skewed by language, with local queries amplifying visibility for regional champions while English queries obscure their presence.