Yijia Fang
2 indexed papers
Publications per year
Top categories
Frequent co-authors
Research Timeline
The paper introduces KBF, a novel black-box auditing protocol that fingerprints LLM APIs by analyzing stable numerical recall near the knowledge boundary, effectively detecting model substitutions and inconsistencies in third-party API endpoints.
The paper introduces KBF, a low-cost black-box auditing protocol that fingerprints LLM APIs by analyzing stable numerical recall near the knowledge boundary, successfully detecting numerous model substitutions and inconsistencies across various production endpoints.
Papers
KBF: Knowledge Boundary as Fingerprint for Language Model and Black-Box API Auditing
The paper introduces KBF, a novel black-box auditing protocol that fingerprints LLM APIs by analyzing stable numerical recall near the knowledge boundary, effectively detecting model substitutions and…