Adds the social profile (callers, communities, timing)
Model Zero v0.43
Adds the social profile: how many accounts and independent communities mention the coin, how reliable those sources have been on earlier calls, whether the talk started before the move, paid promotion, and spam-like repetition.
Completed
Definition
Technical details
Artifact
models/v0.43/model.json
sha256 57a6de02c5dafe5a…
Dataset & compute
Dataset
Solana Launch Dataset v34 (f11)
0 snapshots · 0 launches · sha256 e3b0c44298fc
Compute
Model Zero training server (CPU)
Benchmarks
held-out test set · change vs v0.42
Spotting crashes (−70% within 1h)
8 tests
not enough data yet
Spotting survivors (still traded after 1h)
8 tests
not enough data yet
Spotting 2× runs (within 1h)
8 tests
not enough data yet
Calling the 1h direction (down / flat / up)
8 tests
not enough data yet
Spotting creator dumps (within 1h)
7 tests
not enough data yet
Spotting crashes (within 6h)
8 tests
not enough data yet
Spotting survivors (6h)
8 tests
not enough data yet
Validation metrics (during training)
Signals this version discovered
No candidate improved predictions enough to keep.
0 candidate signals tested on held-out folds of the training launches. Cross-validated AUC → .
What this version relies on
held-out validation launches
This version was trained before Model Zero started measuring which signals each model relies on.
For each signal, how much the score drops when that signal is scrambled across launches. Model Zero learns these weights itself from the data; nobody hand-codes them.
Training runs
Attempt 1Succeeded
08 Oct 14:35 · 00:00:00 · — · $0.00
Timeline
08 Oct 14:35
Model Zero v0.43 not deployed; v0.40 stays live
Reason: no shared benchmarks yet; keeping the live model.
08 Oct 14:35
Benchmark completed: Model Zero v0.43
No locked test set yet (not enough held-out launches with observed outcomes). Benchmarks will be computed when it is locked.
08 Oct 14:35
Benchmark started: Model Zero v0.43
08 Oct 14:35
Model Zero v0.43 searched for new signals; none helped enough to keep
Every candidate was tested on held-out folds of the training launches.
08 Oct 14:35
Training completed: Model Zero v0.43
00:00:00 · trained on 0 examples · $0 compute cost
08 Oct 14:35
Training started: Model Zero v0.43
Solana Launch Dataset v34 (f11) · boosted decision trees · Model Zero training server (CPU)
08 Oct 14:35
Model Zero v0.43 queued
Adds the social profile: how many accounts and independent communities mention the coin, how reliable those sources have been on earlier calls, whether the talk started before the move, paid promotion, and spam-like repetition.