TK Lin · AI citation report

Queries were sent through Perplexity Agent API（web_search 開啟）; the search backend was Perplexity 搜尋（非 Perplexity 的四個模型也用這個搜尋）. Measured: 2026-10-03 17:52（原報告未標示時區）.

All five models used the same search backend, so this measures whether that backend finds the page — not what ChatGPT, Claude or Gemini find with their own native search in their own apps. Numbers may differ from what you see there.

40 of 100 planned questions were completed; 500 records, 197 successful. Unfinished questions are excluded and not extrapolated.

已完成：身分 20 題、公司 20 題。未完成：專長、和心村、個人 IDA 共 60 題。 Not yet translated

Retrieved = the AI's search results include the domain.
Cited = the answer text itself writes out the domain (stricter; styles differ by vendor).

Not measured here: AI crawler visits, trends over time, or sentence-by-sentence accuracy of what AI says.