Researchllmranking manipulationshadow modelsreasoning augmentation
Researchers Demonstrate CORE Manipulates LLM-Based Search Rankings
10.0
Researchers published a study demonstrating CORE, an optimization method that systematically influences LLM-based search rankings, testing Claude-4, Gemini-2.5, GPT-4o and Grok-3 via API. Query-based CORE achieved roughly 77–82% Top-1 promotion while shadow-model approaches (with Llama-3.1-8B proxy) showed lower but transferable gains, and reasoning- versus review-based augmentations affected models differently.
Key Points
- 1Demonstrate CORE boosts target rankings across Claude-4, GPT-4o, Gemini-2.5, and Grok-3, achieving ~77–82%.
- 2Use query-based and shadow-model reverse-engineering, showing optimizations transfer even with approximate surrogates.
- 3Indicate reasoning versus review augmentations affect models differently, enabling targeted manipulation strategies for practitioners.
Scoring Rationale
High novelty, broad scope, and direct exploitability; limited by API-only tests and lack of consumer-interface evaluation.
Sources
Public references used for this report.
Practice with real Ad Tech data
90 SQL & Python problems · 15 industry datasets
Used by DS/ML engineers at top companies
Active Search Campaigns by BudgetEasyHigh CPC Clicks & Poor Landing PagesMediumCampaign ROAS by Attribution ModelHard
250 free problems · No credit card
See all Ad Tech problems
