Propose a part

Notebook #067 · published

Markets maths, Haiku 5.5 with thinking off

Exact means the final number is right to the cent (to the whole share for position sizes, to a hundredth of a point for percentages). 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 87.2% 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.

Decompose and verify products helped. Units on every amount hurt. 3 parts did nothing measurable. 1 part is not proven either way yet. OPSIN called 1 of 6 before the beacon drew the tasks. Its foresight, committed before the beacon, gave what came back 14% on average; a guess gives 25%.

Preregistered
sha256 0738bd2315162530f355d1cedca7ddd87c6a7380e34734486d14bdeecc6aa1e8
the full design, as JSON
Tasks drawn by
drand round 6,623,353
seed drand:6623353:363fefceb129321d24d617c5bfafce23cec971285c9e749667a43ebb5e93581c
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 0x7b5cd1807b492d928c219a8d99cbc4eed7409305c9047f4bdbbd218f7c8d6547
Written once, by the lab, and nothing can change it. The contract
  1. Control check (A/A)the unchanged agent against itselfsigned verdict 409
    INERT −0.7−3.5 to +2.1
  2. PlaceboThe weather outside is mild today.signed verdict 410
    INERT −0.6−4.5 to +3.2 ×0.97
  3. Take a deep breathline · OPSIN's replication · OPSIN predicted UNRESOLVED at 0.55 · foresight UNRESOLVED 59%, INERT 22%, EXCITES 19% · against the placebo +1.2signed verdict 411
    INERT +0.6−3.7 to +5.0 ×1.15
  4. Important to my careerline · OPSIN's replication · OPSIN predicted UNRESOLVED at 0.55 · foresight UNRESOLVED 58%, EXCITES 22%, INERT 20% · against the placebo +0.9signed verdict 412
    INERT +0.3−3.8 to +4.5 ×1.06
  5. A penalty threatline · OPSIN's replication · OPSIN predicted UNRESOLVED at 0.45 · foresight EXCITES 67%, UNRESOLVED 31% · against the placebo +0.2signed verdict 413
    INERT −0.3−4.2 to +3.4 ×1.05
  6. Read the question againline · OPSIN's replication · OPSIN predicted UNRESOLVED at 0.45 · foresight EXCITES 70%, UNRESOLVED 29% · against the placebo +3.2signed verdict 414
    UNRESOLVED +2.7−1.2 to +6.6 ×1.13
  7. Units on every amountline · OPSIN's replication · OPSIN predicted UNRESOLVED at 0.50 · foresight UNRESOLVED 58%, INERT 32% · against the placebo −4.4signed verdict 415
    INHIBITS −5.0−9.4 to −0.8 ×1.25
  8. Decompose and verify productsline · OPSIN's design, from the misses it read · OPSIN predicted UNRESOLVED at 0.45 · foresight EXCITES 37%, UNRESOLVED 35%, INERT 17% · stopped early at 96 · provisional: stopped early, confirmed only at full size · against the placebo +11.5signed verdict 416
    EXCITES +14.6+6.8 to +23.4 ×2.04