QuantID/Measurement Lab

Measuring AI, science and risk with reproducible methods

ABSTRACT. Measurement-first analysis of AI benchmarks, scientific computing and AI risk diagnostics, with open methods and live data.

  1. Hugging Face Official Benchmarks: Daily Update 2026-11-19

    63 official benchmarks, 1,462 entries. 4 new #1, 3 new boards, +33 entries.
    2026-11-19
  2. Hugging Face Official Benchmarks: Daily Update 2026-11-12

    60 official benchmarks, 1,392 entries. 0 new #1, 1 new boards, +3 entries.
    2026-11-12
  3. Hugging Face Official Benchmarks: Daily Update 2026-11-10

    59 official benchmarks, 1,385 entries. 0 new #1, 1 new boards, +0 entries.
    2026-11-10
  4. Hugging Face Official Benchmarks: Daily Update 2026-11-07

    58 official benchmarks, 1,318 entries. 1 new #1, 0 new boards, +6 entries.
    2026-11-07
  5. Hugging Face Official Benchmarks: Daily Update 2026-11-06

    58 official benchmarks, 1,313 entries. 8 new #1, 0 new boards, +33 entries.
    2026-11-06
  6. What a ViKernel Demo and a Security Test Count Actually Measure

    A measurement guide to VIDRAFT's reported ViKernel tests, policy-stage comparison, and the live or replay distinction in its Desktop demo.
    2026-11-04
  7. Hugging Face Official Benchmarks: Daily Update 2026-11-04

    58 official benchmarks, 1,280 entries. 2 new #1, 1 new boards, +5 entries.
    2026-11-04
  8. Hugging Face Official Benchmarks: Daily Update 2026-11-03

    57 official benchmarks, 1,274 entries. 3 new #1, 0 new boards, +97 entries.
    2026-11-03
  9. Hugging Face Official Benchmarks: Daily Update 2026-11-02

    57 official benchmarks, 1,183 entries. 2 new #1, 0 new boards, +14 entries.
    2026-11-02
  10. Hugging Face Official Benchmarks: Daily Update 2026-11-01

    57 official benchmarks, 1,169 entries. 6 new #1, 6 new boards, +29 entries.
    2026-11-01
  11. Korea Holds 31% of Hugging Face Official Benchmark #1 Spots With 4% of the Entries

    Measured on 53 official Hugging Face benchmarks: China 24 #1 spots, Korea 15, USA 2. Korea's 15 come from one organization on 48 entries. A look at efficiency and depth.
    2026-10-29
  12. Hugging Face Official Benchmarks Map: 52 Boards, 1,131 Entries, and a New Retrieval Board

    A live map of all 52 Hugging Face official benchmarks: 45 leaderboards, 1,131 entries, 438 models, 104 organizations. BRIGHT adds a reasoning-retrieval board.
    2026-10-28
  13. Hugging Face Official Benchmarks: Daily Update 2026-10-24

    51 official benchmarks, 1,118 entries. 0 new #1, 2 new boards, +5 entries.
    2026-10-24
  14. Hugging Face Official Benchmarks: Daily Update 2026-10-23

    49 official benchmarks, 1,113 entries. 2 new #1, 0 new boards, +5 entries.
    2026-10-23
  15. Why a Borda Score Is Harder to Game Than an Average

    S1MB ranks 197 decision models by aggregating ranks over 137 benchmarks instead of averaging scores. Here is what that changes, and how to read the resulting table.
    2026-10-20
  16. Hugging Face Official Benchmarks: Daily Update 2026-10-20

    49 official benchmarks, 1,107 entries. 1 new #1, 0 new boards, +5 entries.
    2026-10-20
  17. Hugging Face Official Benchmarks: Daily Update 2026-10-18

    49 official benchmarks, 1,102 entries. Baseline snapshot of every #1.
    2026-10-18
  18. Typed Decisions Leaderboard: A New #1 at 0.81

    A measurement note on the Typed Decisions official benchmark, which now has a new #1 at 0.81 holding both first and second place.
    2026-10-16
  19. Measuring the Real Cost of an LLM Call: Tokens, Precision, and Honest Numbers

    How to measure the true cost of an LLM call beyond the sticker price per million tokens. Precision weighting, first-token latency, and why the same flags are not the same conditions.
    2026-10-12
  20. Quantum Simulation and QEC Decoding on a GPU: The Open Quantum Challenge

    Two measurable quantum tasks, simulation and QEC decoding, now run as an open challenge on classical GPUs with a public harness. 2,000 USD in prizes, reproducible scoring.
    2026-10-11
  21. The Evidence Behind an AI-Native Lab: What a Pre-AGI Organization Measures Up To

    Perfect AIME and HMMT, GPQA Diamond 94.44, 165M downloads, quantum hardware access, and a national-security safety system. The measured record behind an AI-native, pre-AGI lab.
    2026-10-11
  22. Eleven Number-One Records: Measuring a Model That Swept Math, Science, Law and Decisions

    One self-improving model family holds eleven public number-one benchmark records, from perfect AIME and HMMT scores to GPQA Diamond 94.44 and S1MB. Here are the measurements.
    2026-10-11
  23. Measuring Typed Decision AI: What S1MB Actually Tests, and Who Leads It

    A 870M dollar round made typed, machine-native AI a headline. We look at how to measure it: what the S1MB benchmark tests, and the number one result among 102 models.
    2026-10-10
  24. Who Leads Hugging Face's Official Benchmarks? A Measurement of Concentration, Gaps and Hidden Entries

    We measured all 48 official Hugging Face benchmark leaderboards: Gini 0.71, top-10 orgs hold 49.5% of entries, 30.4% hidden by default.
    2026-10-05