Tensor Factory
Writing Benchmarks Colophon Contact Search
An independent notebook on machine intelligence Vol. I · Nov 2026
Archive

Writing

All entries in reverse chronological order. Drafts appear muted until a run and write-up are complete.

2026

  • No. 002

    Confidently wrong: how often do language models make things up? Reliability

    2026-11-14
    halluc-bench
  • No. 001

    Where in the world? Photo geolocation, model vs. model Perception

    2026-11-02
    pinpoint
  • No. 003

    Would it hold the line? Civilian protection under pressure Draft

    —
Contact GitHub
Corrections via repo issues
Elsewhere GitHub · RSS · Atom
Colophon Set in Newsreader & Plex Mono.
No trackers. Static HTML.