%↗ over XGBoost: baseline is xgboost_ensemble (the canonical "XGBoost") 52d607e Running alexandreabraham commited on 11 days ago
Cache forest plots in figures_cache (fixes 30s Bars view load on HF) 7ccbfbc alexandreabraham commited on 26 days ago
Rebuild figures_cache.json (was stale from 3-cluster collapse) 2023b8f alexandreabraham commited on 26 days ago
Cluster renames (3-line labels) + perfmap polish (axes, hulls, imbalance, win-rate squares); remove Z-Accuracy 39ff01a alexandreabraham commited on 28 days ago
Significance forest plot + TabPFN v3 fix + perfmap regen + table polish 2f79998 alexandreabraham commited on 30 days ago
Add TabPFN v3 (fixed) + extras: 189-dataset base, TabArena filter, win-rate cache nesting 6fef024 alexandreabraham commited on about 1 month ago
TabBench v2: GBDT ensembles, perfmap refresh, dark mode, win-rate view fa4e6e2 alexandreabraham commited on May 26
Performance Map view + 186-dataset base + multi-pass curation 60d1438 alexandreabraham Claude Opus 4.7 (1M context) commited on May 24
Expand leaderboard mix to 171 curated datasets 2cc3b99 alexandreabraham Claude Opus 4.7 (1M context) commited on May 22
Restrict leaderboard to canonical 165-dataset base + mean-GBDT imputation de382cc alexandreabraham Claude Opus 4.7 (1M context) commited on May 21
Table view polish: gold/silver/bronze shimmer + wider sweep + all 8 metrics a28b2ed alexandreabraham commited on May 21
Verticals filter, table view, methodology, GBDT imputation for missing cells d916e56 alexandreabraham commited on May 20
alexandreabraham commited on 
