PhilBench quizzes 97 LLMs with the 2020 PhilPapers Survey of 1,785 philosophers

pwlot · x · 2026-11-18

PhilBench turns the 2020 PhilPapers Survey — answered by 1,785 professional philosophers — into a benchmark administered to 97 LLMs, probing what philosophical positions models hold and whether they act on them. The author highlights several standout findings in a thread. An interesting attempt to convert philosophy's largest empirical survey into an LLM evaluation benchmark.

Original post →

More from Models

Models channel →