Taiwan
Founder, Semantic Firewall System (SRCP)|語意責任鏈標準創建者|AI 審計與語義治理框架設計者 我致力於建立 AI 語義治理的下一個標準層(S-Layer), 使語言模型具備: 可審計性(Auditability) 可責任化(Responsibility Mapping) 可計算的語義穩定性(Semantic Stability) 跨模型一致性(Cross-Model Consistency) 我提出 SRCP(Semantic Responsibility Chain Protocol), 並構建 語意防火牆系統(Semantic Firewall System), 可量化模型: 幻覺率 主語錯位 語義責任缺失(SCBKR) 語義污染 計算浪費路徑(Compute Waste Path) 目前我的工作重心: 語義責任鏈標準化 AI 審計工具的通用化(SPI / VPI / SCBKR) 跨模型語義一致性框架(S-Layer) 協助決策層理解模型盲區與不可見風險 🇺🇸 English Version I work on defining the next governance layer for AI — the Semantic Layer (S-Layer), introducing a framework where language models gain: Auditability Responsibility mapping Measurable semantic stability Cross-model consistency I created SRCP (Semantic Responsibility Chain Protocol) and the Semantic Firewall System, enabling quantifiable evaluation of: Hallucination profiles Subject misalignment SCBKR responsibility gaps Semantic contamination Compute waste paths Current focus areas: Standardizing the semantic responsibility chain Building universal AI audit indicators (SPI / VPI / SCBKR) Establishing S-Layer semantic consistency across models Supporting decision-makers in understanding model blind spots and risks 沈耀 888π|Semantic Firewall System Founder “唯真存續|幻象歸零”