Hugging Face's logo

Spaces:
evalstate
/
tb21-annotated
Running

App Files Community
tb21-annotated

Commit History

Today

  1. Gemini CLI: thought summaries filed as messages by Harbor's old converter, counted as reasoning (summary only)

    evalstate HF Staff
    Verified signature
    05c21f3

  2. Trace completeness: share of output tokens in the trace; why not (withheld by provider, not kept by harness, lost in conversion, summary only); counted/inferred basis; undercounted

    evalstate HF Staff
    Verified signature
    fe0df2c

  3. Output tokens as shares: tool calls, messages, reasoning shown, reasoning hidden

    evalstate HF Staff
    Verified signature
    00853bc

Yesterday

  1. Cost estimate: only a fit that reproduces recorded costs prices trials; unpriced wording; PR-aware row page (part 2)

    evalstate HF Staff
    Verified signature
    688f290

  2. Cost estimate: only a fit that reproduces recorded costs prices trials; unpriced wording; PR-aware row page

    evalstate HF Staff
    Verified signature
    8cecfac

  3. Our estimate (leads + fallback discounted), output/hidden reasoning bar, cost estimate, regrouped gaps, waffle flags, Harbor Hub trial links (part 5)

    evalstate HF Staff
    Verified signature
    1886473

  4. Our estimate (leads + fallback discounted), output/hidden reasoning bar, cost estimate, regrouped gaps, waffle flags, Harbor Hub trial links (part 4)

    evalstate HF Staff
    Verified signature
    6d725e3

  5. Our estimate (leads + fallback discounted), output/hidden reasoning bar, cost estimate, regrouped gaps, waffle flags, Harbor Hub trial links (part 3)

    evalstate HF Staff
    Verified signature
    bed26c0

  6. Our estimate (leads + fallback discounted), output/hidden reasoning bar, cost estimate, regrouped gaps, waffle flags, Harbor Hub trial links (part 2)

    evalstate HF Staff
    Verified signature
    84ce00f

  7. Our estimate (leads + fallback discounted), output/hidden reasoning bar, cost estimate, regrouped gaps, waffle flags, Harbor Hub trial links

    evalstate HF Staff
    Verified signature
    b8a3c5c

  8. Link the submissions Space

    evalstate HF Staff
    Verified signature
    fd1c34e

  9. Viewer: fold many argument-value fields per call (atif-scan 43ae555) (part 5)

    evalstate HF Staff
    Verified signature
    a63d5f0

  10. Viewer: fold many argument-value fields per call (atif-scan 43ae555) (part 4)

    evalstate HF Staff
    Verified signature
    c807152

  11. Viewer: fold many argument-value fields per call (atif-scan 43ae555) (part 3)

    evalstate HF Staff
    Verified signature
    56a5b0b

  12. Viewer: fold many argument-value fields per call (atif-scan 43ae555) (part 2)

    evalstate HF Staff
    Verified signature
    50e0e7c

  13. Viewer: fold many argument-value fields per call (atif-scan 43ae555)

    evalstate HF Staff
    Verified signature
    77fbc9e

  14. One 'output not in trace' measure on the leaderboard; gaps table grouped; row page two-column layout

    evalstate HF Staff
    Verified signature
    19a8efc

  15. Restore row exports (removed by an over-broad delete pattern) (part 6)

    evalstate HF Staff
    Verified signature
    ac26316

  16. Restore row exports (removed by an over-broad delete pattern) (part 5)

    evalstate HF Staff
    Verified signature
    57ea25f

  17. Restore row exports (removed by an over-broad delete pattern) (part 4)

    evalstate HF Staff
    Verified signature
    dd7b1f5

  18. Restore row exports (removed by an over-broad delete pattern) (part 3)

    evalstate HF Staff
    Verified signature
    a79451b

  19. Restore row exports (removed by an over-broad delete pattern) (part 2)

    evalstate HF Staff
    Verified signature
    4c0c727

  20. Restore row exports (removed by an over-broad delete pattern)

    evalstate HF Staff
    Verified signature
    bb99763

  21. Dense tables back on the home page (leaderboard, gaps & accounting, families, checks × rows); badges reduced to plain coloured text

    evalstate HF Staff
    Verified signature
    65bbcca

  22. Quieter design: leaderboard + row pages + compare; full trajectories for highlighted trials (atif-scan 90c8670 --viewer-split) (part 6)

    evalstate HF Staff
    Verified signature
    9ccc85f

  23. Quieter design: leaderboard + row pages + compare; full trajectories for highlighted trials (atif-scan 90c8670 --viewer-split) (part 5)

    evalstate HF Staff
    Verified signature
    4f9d72f

  24. Quieter design: leaderboard + row pages + compare; full trajectories for highlighted trials (atif-scan 90c8670 --viewer-split) (part 4)

    evalstate HF Staff
    Verified signature
    c782ca8

  25. Quieter design: leaderboard + row pages + compare; full trajectories for highlighted trials (atif-scan 90c8670 --viewer-split) (part 3)

    evalstate HF Staff
    Verified signature
    3220745

  26. Quieter design: leaderboard + row pages + compare; full trajectories for highlighted trials (atif-scan 90c8670 --viewer-split) (part 2)

    evalstate HF Staff
    Verified signature
    8203088

  27. Quieter design: leaderboard + row pages + compare; full trajectories for highlighted trials (atif-scan 90c8670 --viewer-split)

    evalstate HF Staff
    Verified signature
    5d9b93c

  28. Models served/fallback, reasoning and output-token accounting, drill-down to trials, tooltip fix

    evalstate HF Staff
    Verified signature
    8c073f5

  29. Data holes & accounting; leaderboard rows scored as listed (atif-scan v0.8.17-2-g4fbd8b8)

    evalstate HF Staff
    Verified signature
    2cbfd05

  30. TB2.1 annotated: 22 leaderboard rows, atif-scan v0.8.17 highlights and trajectories

    evalstate HF Staff
    Verified signature
    edc735f

  31. initial commit

    evalstate HF Staff
    Verified signature
    bdcde10