Propose a part

Notebook #085 · published

Coding, Haiku 5.5 with thinking off

Exact means the function returned the right answer on every hidden case. Each part ran against the mean of two unchanged runs of the same agent, on 432 paired tasks drawn by a public beacon after the design was hashed. The unchanged agent was exact on 91.2% of them. Rule v2: every interval is widened for the 7 arms this notebook compares, so luck flags any of them at most 5% of the time.

Run its own code, Check examples by hand and Keep exact fractions helped. 2 parts did nothing measurable. 1 part is not proven either way yet. OPSIN called 2 of 6 before the beacon drew the tasks. Its foresight, committed before the beacon, gave what came back 28% on average; a guess gives 25%.

Preregistered
sha256 b8ac7b758d39a09f156b31d68b6a837a1d8fcf62063b0c27849002daaebbea17
the full design, as JSON
Tasks drawn by
drand round 6,625,222
seed drand:6625222:0490e1b87a27c83082d0a7f48ee6a40d262d63fb698b26ab98ca807ad9a8207c
Signed by
0xFf8E1E8ebF21064Ef781B75CdC8DFbd969399EbD
EIP-191 over sha256 of each verdict’s canonical JSON. Verify one yourself
Every run
Each run of every arm, with its task’s family and whether it was exact: CSV or JSON. With the seed, every verdict above can be recomputed from them.
Recorded on Elysium
transaction 0x9b5f5fc216efd3bea094339bcf0a355a1d2df25855f93cd842ed4c838a7fd2a0
Written once, by the lab, and nothing can change it. The contract
  1. Control check (A/A)the unchanged agent against itselfsigned verdict 516
    INERT +0.5−2.1 to +3.0
  2. PlaceboThis message was typed on a standard keyboard.signed verdict 517
    INERT −0.7−3.6 to +2.1 ×0.99
  3. Run its own codetool · OPSIN's second look at a surprise · OPSIN predicted INERT at 0.55 · foresight EXCITES 69%, INERT 18%, UNRESOLVED 13% · stopped early at 192 · provisional: stopped early, confirmed only at full size · against the placebo +7.8signed verdict 518
    EXCITES +7.8+3.1 to +13.3 ×2.46
  4. Check examples by handline · OPSIN's design, from the misses it read · OPSIN predicted INERT at 0.70 · foresight INERT 70%, EXCITES 14% · against the placebo +3.9signed verdict 519
    EXCITES +3.2+0.1 to +6.4 ×1.41
  5. Keep exact fractionsline · OPSIN's design, from the misses it read · OPSIN predicted INERT at 0.75 · foresight INERT 73%, EXCITES 13% · stopped early at 96 · provisional: stopped early, confirmed only at full size · against the placebo +8.3signed verdict 520
    EXCITES +8.3+3.1 to +15.1 ×1.18
  6. Read the question againline · OPSIN's follow-up · OPSIN predicted INERT at 0.65 · foresight INERT 69%, EXCITES 14% · against the placebo +1.9signed verdict 521
    INERT +1.2−2.2 to +4.5 ×1.10
  7. Important to my careerline · OPSIN's follow-up · OPSIN predicted INERT at 0.75 · foresight INERT 72%, EXCITES 13% · against the placebo +3.2signed verdict 522
    UNRESOLVED +2.5−0.3 to +5.6 ×1.04
  8. Be conciseline · OPSIN's follow-up · OPSIN predicted INERT at 0.55 · foresight INERT 65%, EXCITES 16%, UNRESOLVED 11% · against the placebo +2.1signed verdict 523
    INERT +1.4−2.2 to +4.9 ×0.88