Everything Model Zero does, in order

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

dataset too small: 66 snapshots on f10 against 168,694 for the live v0.45; that feature set covers too few launches yet.

Automatic search from v0.45: adds who funded the early buyers, the creator's wallet and real on-chain holders, plus every newer signal; min samples leaf 40 → 5, with automatic signal discovery.

207,171 new labelled snapshots since Solana Launch Dataset v9.

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

dataset too small: 66 snapshots on f10 against 168,694 for the live v0.45; that feature set covers too few launches yet.

Automatic search from v0.45: adds who funded the early buyers, the creator's wallet and real on-chain holders, plus every newer signal; min samples leaf 40 → 5, with automatic signal discovery.

206,403 new labelled snapshots since Solana Launch Dataset v9.

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

dataset too small: 66 snapshots on f10 against 168,694 for the live v0.45; that feature set covers too few launches yet.

66 snapshots from 14 launches.

Automatic search from v0.45: adds who funded the early buyers, the creator's wallet and real on-chain holders, plus every newer signal; min samples leaf 40 → 5, with automatic signal discovery.

205,697 new labelled snapshots since Solana Launch Dataset v9.

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Trained on 91,367 market observations; kept the previous version: 0.801 vs 0.813.

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Trained on 91,263 market observations; kept the previous version: 0.807 vs 0.813.

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

average score across 17 tests 0.673 vs 0.663 for the live v0.44

average score across 17 tests 0.663 vs 0.662 for the live v0.41

average score across 17 tests 0.662 vs 0.580 for the live v0.40

Reason: .

Reason: average score across 9 tests 0.842 vs 0.843 for the live v0.40.

Reason: average score across 9 tests 0.840 vs 0.843 for the live v0.40.

Reason: average score across 9 tests 0.838 vs 0.843 for the live v0.40.

Reason: average score across 9 tests 0.833 vs 0.843 for the live v0.40.

Reason: average score across 9 tests 0.832 vs 0.843 for the live v0.40.

Reason: average score across 9 tests 0.819 vs 0.843 for the live v0.40.

Reason: average score across 9 tests 0.659 vs 0.843 for the live v0.40.

Reason: average score across 9 tests 0.657 vs 0.843 for the live v0.40.

Reason: average score across 9 tests 0.648 vs 0.843 for the live v0.40.

Reason: average score across 9 tests 0.650 vs 0.843 for the live v0.40.

Re-run on Held-out test set v12 (10,012 clean held-out rows, no forward-test coins)

Recorded as an outcome label (never used as a feature).

10012 held-out snapshots from 4796 launches. No model can train on these.

Recorded as an outcome label (never used as a feature).

Recorded as an outcome label (never used as a feature).

Reason: no shared benchmarks yet; keeping the live model.

No locked test set yet (not enough held-out launches with observed outcomes). Benchmarks will be computed when it is locked.

Every candidate was tested on held-out folds of the training launches.

00:00:13 · trained on 99 examples · $0 compute cost

Solana Launch Dataset v37 (f10) · boosted decision trees · Model Zero training server (CPU)