Per-track results
Benchmark Tracks
wbc/279f6f6ecb9d
anthropometrics, nightingale
Scored split: test. Primary metric: r2 (higher is better).
| # | Predictor | Method | Absolute test R² | Test n |
|---|
| 1 | HPP Curated Panel | TabFM | 0.2106 | 438 |
| 2 | HPP Curated Panel | TabPFN 3.5 | 0.2088 | 438 |
| 3 | HPP Curated Panel | TabICL | 0.2066 | 438 |
| 4 | HPP Curated Panel | Ridge | 0.1976 | 438 |
| 5 | HPP Curated Panel | TabDPT | 0.1974 | 438 |
| 6 | HPP Curated Panel | TabPFN v2 | 0.1855 | 438 |
| 7 | HPP Curated Panel | TabSwift | 0.1813 | 438 |
| 8 | HPP Curated Panel | RealMLP | 0.1682 | 438 |
| 9 | HPP Curated Panel | Gradient-boosted trees | 0.1368 | 438 |
wbc/3050268d0831
anthropometrics, cgm
Scored split: test. Primary metric: r2 (higher is better).
| # | Predictor | Method | Absolute test R² | Test n |
|---|
| 1 | Participant Embedding | Ridge | 0.01129 | 593 |
wbc/424ea3ca2dd3
anthropometrics, dxa
Scored split: test. Primary metric: r2 (higher is better).
| # | Predictor | Method | Absolute test R² | Test n |
|---|
| 1 | HPP Curated Panel | TabPFN 3.5 | 0.04137 | 749 |
| 2 | HPP Curated Panel | TabICL | 0.03796 | 749 |
| 3 | HPP Curated Panel | TabDPT | 0.03792 | 749 |
| 4 | HPP Curated Panel | TabSwift | 0.03759 | 749 |
| 5 | HPP Curated Panel | RealMLP | 0.03757 | 749 |
| 6 | HPP Curated Panel | TabFM | 0.03728 | 749 |
| 7 | HPP Curated Panel | Ridge | 0.03596 | 749 |
| 8 | HPP Curated Panel | TabPFN v2 | 0.03548 | 749 |
| 9 | HPP Curated Panel | Gradient-boosted trees | 0.02442 | 749 |
wbc/6e2f00e0e15c
anthropometrics, watchpat_actigraphy, watchpat_hr, watchpat_pat, watchpat_spo2
Scored split: test. Primary metric: r2 (higher is better).
| # | Predictor | Method | Absolute test R² | Test n |
|---|
| 1 | Participant Embedding | Ridge | 0.03902 | 228 |
wbc/bc931be17c12
anthropometrics, dxa
Scored split: test. Primary metric: r2 (higher is better).
| # | Predictor | Method | Absolute test R² | Test n |
|---|
| 1 | HPP Curated Panel | TabICL | 0.08629 | 687 |
| 2 | HPP Curated Panel | TabDPT | 0.0853 | 687 |
| 3 | HPP Curated Panel | TabSwift | 0.08472 | 687 |
| 4 | HPP Curated Panel | TabFM | 0.08449 | 687 |
| 5 | HPP Curated Panel | TabPFN 3.5 | 0.0802 | 687 |
| 6 | HPP Curated Panel | TabPFN v2 | 0.07868 | 687 |
| 7 | HPP Curated Panel | RealMLP | 0.07721 | 687 |
| 8 | HPP Curated Panel | Ridge | 0.07558 | 687 |
| 9 | HPP Curated Panel | Gradient-boosted trees | 0.07556 | 687 |
wbc/d3d91ac6bb48
anthropometrics, nightingale
Scored split: test. Primary metric: r2 (higher is better).
| # | Predictor | Method | Absolute test R² | Test n |
|---|
| 1 | HPP Curated Panel | TabFM | 0.1019 | 696 |
| 2 | HPP Curated Panel | TabPFN 3.5 | 0.09665 | 696 |
| 3 | HPP Curated Panel | TabICL | 0.09465 | 696 |
| 4 | HPP Curated Panel | TabPFN v2 | 0.08763 | 696 |
| 5 | HPP Curated Panel | TabDPT | 0.08656 | 696 |
| 6 | HPP Curated Panel | TabSwift | 0.08403 | 696 |
| 7 | HPP Curated Panel | Ridge | 0.08202 | 696 |
| 8 | HPP Curated Panel | RealMLP | 0.06695 | 696 |
| 9 | HPP Curated Panel | Gradient-boosted trees | 0.05879 | 696 |
wbc/fd32ee0a0d27
anthropometrics, ecg_signal
Scored split: test. Primary metric: r2 (higher is better).
| # | Predictor | Method | Absolute test R² | Test n |
|---|
| 1 | Participant Embedding | Ridge | 0.06324 | 271 |