MUMBAI, India, Sept. 28 -- Intellectual Property India has published a patent application (202641060131 A) filed by Navin Senguttuvan; and Prashanthie Velmurugan on May 12, 2026, for Weighted Cognitive State Vector For Llm Prompt Optimization.
Inventors include Navin Senguttuvan; and Prashanthie Velmurugan.
The application for the patent was published on September 25, 2026, under issue no. 39/2026.
Abstract: The present invention relates to a system and method for optimizing prompts for large language models using a Weighted Cognitive State Vector (WCSV). The system extracts privacy-preserving features from user-generated inputs and maps them to a multi-dimensional cognitive state representation. The state is updated using dimension-specific temporal decay and used to dynamically generate weighted prompt components. The invention enables adaptive, context-aware prompt generation while preserving user privacy by avoiding transmission of raw data. The system supports multi-scale temporal modeling and a tiered privacy architecture, thereby improving personalization and efficiency in language model interactions.
Disclaimer: Curated by HT Syndication.
