The Language of the Question Selects the Market: Query Language and Exit IP as Separable Factors in Commercial Recommendations from a Generative Search Interface

arXiv:2608.30052 · cs.IR, cs.CL, cs.CY · Submitted 2026-08-30 · Read on arXiv

cs.IR, cs.CL, cs.CY

Submitted: 2026-08-30

Updated: 2026-08-30

Comments: 13 figures-free pages, 6 tables. Data, collectors and 246 run records: Zenodo DOI 10.5281/zenodo.22181306

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Alignment — 4× in this paper · explained in 1 episode
• Benchmark — 3× in this paper · explained in 2 episodes
• Personalization — 2× in this paper · explained in 2 episodes
• Generative Recommendation — 1× in this paper · explained in 2 episodes
• Provenance — 1× in this paper · explained in 3 episodes
• Robustness — 1× in this paper · explained in 5 episodes

Sources

• Non-Determinism of "Deterministic" LLM Settings
• Retrieval-augmented generation in multilingual settings
• "Global is Good, Local is Bad?": Understanding Brand Bias in LLMs
• Multilingual Retrieval Augmented Generation for Culturally-Sensitive Tasks: A Benchmark for Cross-lingual Robustness
• Large Language Models as Recommender Systems: A Study of Popularity Bias
• Global MMLU: Understanding and Addressing Cultural and Linguistic Biases in Multilingual Evaluation
• Not All Countries Celebrate Thanksgiving: On the Cultural Dominance in Large Language Models
• Not All Languages are Equal: Insights into Multilingual Retrieval-Augmented Generation
• MMLU-ProX: A Multilingual Benchmark for Advanced Large Language Model Evaluation
• How Large Language Models Source Brand Reputation Across Languages and Markets
• The Language Blind Spot: How Query Language and Brand Recognition Tier Shape AI-Constructed Brand Reputation Across Twelve European Languages
• Where Does the Noise Come From? A Variance-Components Decomposition of Non-Determinism in LLM Brand Answers

Related papers

• The Price of Isolation: Estimating the Ecosystem Cost of Symmetric Two-Sided A/B Testing
• SCAR: Semantic Continuity-Aware Retrieval for Efficient Context Expansion in RAG
• MixLoRA-DSI: Dynamically Expandable Mixture-of-LoRA Experts for Rehearsal-Free Generative Retrieval over Dynamic Corpora
• RRCM: Ranking-Driven Retrieval over Collaborative and Meta Memories for LLM Recommendation
• Right Family, Wrong Skill: Evaluating Risk Exposure in Agent Skill Retrieval
• UltRAG: a Universal Simple Scalable Recipe for Knowledge Graph RAG