This page summarizes the public data currently exposed through the Results Browser.
The current HDLSS browser bundle exposes the refreshed local slice: 57,217 run rows across 278 profiles and 70 datasets, including the Val-21 phase-2 RV holdout.
Those rows are surfaced in the browser as four linked tables: dataset summaries, profile summaries, dataset/profile aggregates, and the underlying per-seed run explorer.
Campaign Scope Filter
The browser scope filter is driven by campaign-level dataset coverage. Broad-panel campaigns stay in the full view, while diagnostic-only campaigns stay in the diagnostic view.
| Filter | Meaning |
|---|---|
| All campaigns | Every campaign/profile slice currently exposed by the public HDLSS benchmark bundle. |
| Full benchmark panel | Profiles evaluated across the broad public benchmark panel (typically 62-63 datasets). |
| Diagnostic subset | Profiles evaluated on the 24-dataset diagnostic subset used for narrower probes and weighting checks. |
Val-20 Pipeline remains in the full-panel group because the campaign as a whole publishes against the broad benchmark surface, even though some profiles are narrower probes.
Campaigns
| Campaign | Code | Scope | Profiles | Datasets | Families | What it covers |
|---|---|---|---|---|---|---|
| Val-18 Anchor | val18_anchors | Full benchmark panel | 8 | 63 | Anchor | Baseline and bypass anchor controls across the broad benchmark panel. |
| Val-18 Classifier | val18_classifiers | Full benchmark panel | 40 | 63 | Clf Only, Pool, Stage | Classifier-pool and classifier-ablation runs across the broad benchmark panel. |
| Val-18 Clf Oracle | val18_cls_oracle_wt | Diagnostic subset | 9 | 24 | Clf Oracle | Classifier-oracle weighting diagnostics on the diagnostic subset. |
| Val-18 FS | val18_singletons | Full benchmark panel | 78 | 63 | FS RAW, FS SCAF | Raw and scaffolded feature-selection singletons across the broad benchmark panel. |
| Val-18 MNPO | val18_mnpo | Full benchmark panel | 23 | 63 | Oracle | MNPO oracle weighting and ablation runs across the broad benchmark panel. |
| Val-18 Stage | val18_stage | Diagnostic subset | 37 | 24 | DistFit, Stage, V18 Prefilter | Pipeline stage-ordering probes on the diagnostic subset. |
| Val-19 Bridge | val19_classifiers | Full benchmark panel | 13 | 63 | Clf Only, V19 Bridge | Val-19 bridge specialists and classifier promotions across the broad benchmark panel. |
| Val-19 Random FS | val19_random_fs | Full benchmark panel | 2 | 63 | FS RAW, FS SCAF | Randomized feature-selection controls across the broad benchmark panel. |
| Val-20 Bridge | val20_bridge | Full benchmark panel | 6 | 63 | V20 Bridge | Val-20 bridge reruns that align anchor and promotion candidates on the broad benchmark panel. |
| Val-20 Ensemble | val20_ensemble | Diagnostic subset | 13 | 24 | V20 Ensemble, V20 Oracle Diag | Diagnostic-subset ensemble extraction and voting probes. |
| Val-20 FLAML | val20_flaml | Diagnostic subset | 22 | 24 | V20 FLAML | Diagnostic-subset FLAML and frontier tuning probes. |
| Val-20 LR Mitigation | val20_lr | Full benchmark panel | 3 | 62 | V20 LR Mitigation | Broad-panel logistic-regression overselection countermeasures. |
| Val-20 Pipeline | val20_pipeline | Full benchmark panel | 6 | 63 | V18 Prefilter | Pipeline simplification probes that span both the broad benchmark panel and the diagnostic subset. |
| Val-20 TabPFN | val20_tabpfn | Full benchmark panel | 3 | 63 | V20 TabPFN | Broad-panel TabPFN gating checks. |
| Val-20 Tune-First | val20_tune_first | Diagnostic subset | 8 | 24 | V20 Tune-First | Diagnostic-subset tune-first classifier-selection probes. |
| Val-21 Phase 1 | val21_phase1 | Full benchmark panel | 7 | 63 | Val-21 Current Default, Val-21 Integrated, Val-21 Winner, Val-21 Winner Lockout, Val-21 Winner Meta-Selector, Val-21 Winner Prefilter, Val-21 Winner Regime | Public browser slice for this campaign. |
| Val-21 Phase 2 | val21_phase2 | Diagnostic subset | 3 | 8 | Val-21 Current Default, Val-21 Winner, Val-21 Winner Lockout | Public browser slice for this campaign. |
Families
| Family | Code | Profiles | Datasets | Scope | Campaigns | What it isolates |
|---|---|---|---|---|---|---|
| Anchor | A | 8 | 63 | Full benchmark panel | Val-18 Anchor | Baseline and bypass anchor controls. |
| Clf Only | C_ONLY | 30 | 24 | Full benchmark panel | Val-18 Classifier, Val-19 Bridge | Single-classifier bridge and specialist promotions. |
| Clf Oracle | W | 9 | 24 | Diagnostic subset | Val-18 Clf Oracle | Classifier-oracle reweighting diagnostics. |
| DistFit | D | 19 | 24 | Diagnostic subset | Val-18 Stage | Distribution-fitting and CDF-transform variants. |
| FS RAW | M_RAW | 40 | 63 | Full benchmark panel | Val-18 FS, Val-19 Random FS | Raw feature-selection singleton runs. |
| FS SCAF | M_SCAFFOLD | 40 | 63 | Full benchmark panel | Val-18 FS, Val-19 Random FS | Scaffolded feature-selection singleton runs. |
| Oracle | N | 23 | 63 | Full benchmark panel | Val-18 MNPO | MNPO oracle weighting and ablation profiles. |
| Pool | C | 12 | 63 | Full benchmark panel | Val-18 Classifier | Classifier-pool comparisons and ensemble candidate pools. |
| Stage | S | 7 | 24 | Full benchmark panel, Diagnostic subset | Val-18 Classifier, Val-18 Stage | Stage-ordering probes. |
| V18 Prefilter | P | 22 | 63 | Full benchmark panel, Diagnostic subset | Val-18 Stage, Val-20 Pipeline | Prefilter and dimensionality-folding probes. |
| V19 Bridge | V | 6 | 63 | Full benchmark panel | Val-19 Bridge | Val-19 bridge reruns and promotions. |
| V20 Bridge | B | 6 | 63 | Full benchmark panel | Val-20 Bridge | Public browser family in the benchmark bundle. |
| V20 Ensemble | E | 7 | 24 | Diagnostic subset | Val-20 Ensemble | Val-20 ensemble extraction and voting variants. |
| V20 FLAML | F | 22 | 24 | Diagnostic subset | Val-20 FLAML | Val-20 FLAML and tuned promotion probes. |
| V20 LR Mitigation | L | 3 | 62 | Full benchmark panel | Val-20 LR Mitigation | Val-20 logistic-regression mitigation profiles. |
| V20 Oracle Diag | O | 6 | 24 | Diagnostic subset | Val-20 Ensemble | Public browser family in the benchmark bundle. |
| V20 TabPFN | T | 3 | 63 | Full benchmark panel | Val-20 TabPFN | Val-20 TabPFN gating checks. |
| V20 Tune-First | TF | 8 | 24 | Diagnostic subset | Val-20 Tune-First | Val-20 tune-first bridge profiles. |
| Val-21 Current Default | V21_CURRENT_DEFAULT | 1 | 66 | Full benchmark panel, Diagnostic subset | Val-21 Phase 1, Val-21 Phase 2 | Public browser family in the benchmark bundle. |
| Val-21 Integrated | V21_INTEGRATED | 1 | 63 | Full benchmark panel | Val-21 Phase 1 | Public browser family in the benchmark bundle. |
| Val-21 Winner | V21_WINNER | 1 | 66 | Full benchmark panel, Diagnostic subset | Val-21 Phase 1, Val-21 Phase 2 | Public browser family in the benchmark bundle. |
| Val-21 Winner Lockout | V21_WINNER_LOCKOUT | 1 | 66 | Full benchmark panel, Diagnostic subset | Val-21 Phase 1, Val-21 Phase 2 | Public browser family in the benchmark bundle. |
| Val-21 Winner Meta-Selector | V21_WINNER_METASEL | 1 | 63 | Full benchmark panel | Val-21 Phase 1 | Public browser family in the benchmark bundle. |
| Val-21 Winner Prefilter | V21_WINNER_PREFILTER | 1 | 63 | Full benchmark panel | Val-21 Phase 1 | Public browser family in the benchmark bundle. |
| Val-21 Winner Regime | V21_WINNER_REGIME | 1 | 63 | Full benchmark panel | Val-21 Phase 1 | Public browser family in the benchmark bundle. |
Datasets
| Dataset | Tier | Domain | Platform | Samples / Features | Current top profile | SOTA band |
|---|---|---|---|---|---|---|
11-Tumor (Su) (tumor11_su) | Hard | genomics | Affy HG-U133A | 174 / 12,533 | N01_core_legacy_voting | 0.55-0.80 (above) |
9-Tumors (OpenML) (tumor9_openml) | Hard | genomics | cDNA | 60 / 5,726 | V19_C03_old_regime_mnpo_full64 | 0.45-0.65 (above) |
ARCENE (NIPS 2003 FS Challenge) (arcene_nips03) | Medium | non_genomic | cDNA | 200 / 10,000 | P12_fold_tensor_sketch | 0.75-0.89 (above) |
BASEHOCK Text (Scikit-feature) (rv_basehock) | Easy | genomics | cDNA | 1,993 / 4,862 | P14_lowpn_fast_filter | 0.90-0.97 (within) |
Brain Tumor 2 — Nutt 2003 (50 x 12625, 4-class) (brain_tumor_2_50_4class) | Medium | genomics | cDNA | 50 / 12,625 | V19_C03_old_regime_mnpo_full64 | 0.86-0.95 (above) |
Breast Cancer Gene Expression (HF: mubashir1837) (hf_breast_ge_mubashir1837) | Hard | genomics | cDNA | 51 / 28,278 | A01_simple_anchor_after_fs | 0.50-0.70 (above) |
Breast Cancer Prognosis (van ‘t Veer) (breast_vantveer) | Medium | genomics | cDNA | 97 / 24,481 | N07_core_banzhaf_no_payoff | 0.75-0.87 (within) |
Carcinom 11-class (Scikit-feature) (carcinom_11class) | Hard | genomics | cDNA | 174 / 9,182 | N32_core_banzhaf_perf_complex_stab | 0.72-0.88 (above) |
CLL_SUB_111 (Scikit-feature) (cll_sub_111) | Medium | genomics | cDNA | 111 / 11,340 | N41_core_banzhaf_perf_only_5x5 | 0.78-0.90 (above) |
CNS / Brain Tumors (Pomeroy) (cns_pomeroy) | Medium | genomics | cDNA | 60 / 7,129 | M_SCAFFOLD_boruta | 0.74-0.86 (within) |
COIL20 Objects (Scikit-feature) (rv_coil20) | Medium | genomics | cDNA | 1,440 / 1,024 | V21_CURRENT_DEFAULT | 0.85-0.95 (above) |
Colon Cancer (Alon) (colon_alon) | Medium | genomics | Affy HG-U133A | 62 / 2,000 | C02_pool_legacy_plus_tabpfn | 0.80-0.93 (above) |
CuMiDa Brain Subtypes (GSE50161) (cumida_brain_gse50161) | Hard | genomics | cDNA | 108 / 54,675 | A01_simple_anchor_after_fs | 0.88-0.98 (above) |
CuMiDa Breast Subtypes (GSE45827) (cumida_breast_gse45827) | Hard | genomics | cDNA | 151 / 54,675 | A02_default_anchor_after_fs | 0.82-0.95 (above) |
CuMiDa Colorectal (GSE44861) (cumida_colorectal_gse44861) | Medium | genomics | cDNA | 111 / 22,277 | A01_simple_anchor_after_fs | 0.88-0.96 (within) |
CuMiDa Gastric (GSE54129) (cumida_gastric_gse54129) | Easy | genomics | cDNA | 132 / 54,675 | A01_simple_anchor_after_fs | 0.90-0.98 (above) |
CuMiDa Head/Neck (GSE12452) (cumida_headneck_gse12452) | Medium | genomics | cDNA | 41 / 54,675 | A02_default_anchor_after_fs | 0.78-0.90 (above) |
CuMiDa Leukemia Subtypes (cumida_leukemia_subtypes) | Very Hard | genomics | cDNA | 281 / 22,283 | A02_default_anchor_after_fs | 0.85-0.97 (within) |
CuMiDa Lung (GSE19804) (cumida_lung_gse19804) | Medium | genomics | cDNA | 120 / 54,675 | A07_ref_before_fs | 0.85-0.95 (above) |
CuMiDa Ovarian (GSE26712) (cumida_ovarian_gse26712) | Medium | genomics | cDNA | 195 / 22,283 | A01_simple_anchor_after_fs | 0.90-0.98 (above) |
CuMiDa Pancreatic (GSE16515) (cumida_pancreatic_gse16515) | Medium | genomics | cDNA | 52 / 54,613 | A05_skip_df_ref | 0.80-0.92 (above) |
CuMiDa Prostate (GSE6919) (cumida_prostate_gse6919) | Medium | genomics | cDNA | 171 / 12,625 | V20_L02_diversity_top3_full64 | 0.62-0.78 (above) |
CuMiDa Renal (GSE53757) (cumida_renal_gse53757) | Medium | genomics | cDNA | 144 / 54,675 | A03_ref_anchor_after_fs | 0.80-0.92 (above) |
DEXTER (NIPS 2003 FS Challenge) (dexter_nips03) | Medium | non_genomic | cDNA | 600 / 20,000 | N30_core_banzhaf_perf_only | 0.75-0.89 (above) |
DLBCL (Shipp) (dlbcl_shipp) | Easy | genomics | Affy HG-U95 | 77 / 5,469 | A02_default_anchor_after_fs | 0.90-0.98 (above) |
DOROTHEA (NIPS 2003 FS Challenge) (dorothea_nips03) | Hard | non_genomic | cDNA | 1,150 / 10,000 | D07_df_cv_after | 0.50-0.74 (above) |
GCM (Ramaswamy) (gcm_ramaswamy) | Hard | genomics | Affy HG-U133A | 190 / 16,063 | D18_multimodal_rank | 0.50-0.65 (above) |
GISETTE (NIPS 2003 FS Challenge) (gisette_nips03) | Medium | non_genomic | cDNA | 7,000 / 5,000 | P15_lowpn_all_features | 0.75-0.89 (above) |
GLA-BRA-180 (Scikit-feature) (gla_bra_180) | Hard | genomics | cDNA | 180 / 49,151 | V19_C03_old_regime_mnpo_full64 | 0.58-0.72 (above) |
Glioma (50 x 4434, 4-class) (glioma_50_4class) | Medium | genomics | cDNA | 50 / 4,434 | N11_core_banzhaf_cvar | 0.88-0.97 (within) |
Glioma (GLI_85) (gli_85) | Medium | genomics | Affy HG-U133A | 85 / 22,283 | A05_skip_df_ref | 0.75-0.87 (above) |
ISOLET Audio (Scikit-feature) (rv_isolet) | Hard | genomics | cDNA | 1,560 / 617 | V21_CURRENT_DEFAULT | 0.60-0.78 (above) |
Leukemia (Golub) (leukemia_golub) | Easy | genomics | Affy HG-U95 | 72 / 7,129 | A01_simple_anchor_after_fs | 0.93-1.00 (within) |
Lung Cancer (OpenML Lung, 5-class) (lung_gordon) | Hard | genomics | Affy HG-U133A | 203 / 12,600 | A08_ref_no_regime_gating | 0.75-0.88 (above) |
Lymphoma-11 (OpenML) (lymphoma_11) | Hard | genomics | Affy HG-U133A | 96 / 4,026 | V21_WINNER_REGIME | 0.45-0.65 (above) |
Lymphoma-3 (OpenML) (lymphoma_3) | Medium | genomics | Affy HG-U133A | 66 / 4,026 | A01_simple_anchor_after_fs | 0.80-0.90 (above) |
Lymphoma-9 (OpenML) (lymphoma_9) | Hard | genomics | Affy HG-U133A | 96 / 4,026 | V21_WINNER_REGIME | 0.50-0.70 (above) |
MADELON (NIPS 2003 FS Challenge) (madelon_nips03) | Medium | non_genomic | cDNA | 2,600 / 500 | C10_only_tabpfn_full64 | 0.75-0.89 (within) |
MLL (OpenML) (mll_microarray) | Easy | genomics | Affy HG-U95 | 72 / 12,582 | A02_default_anchor_after_fs | 0.88-0.96 (above) |
MLL Leukemia — Armstrong 2002 (72 x 12533, 3-class) (leukemia_1_72_3class) | Medium | genomics | cDNA | 72 / 12,533 | A01_simple_anchor_after_fs | 0.85-0.96 (above) |
NCI (61 x 5244, 8-class proxy) (nci_61_8class) | Hard | genomics | cDNA | 61 / 5,244 | A07_ref_before_fs | 0.50-0.68 (above) |
NCI60 (Ross) (nci60_ross) | Hard | genomics | Affy HG-U95 | 64 / 6,830 | M_RAW_chi_square | 0.40-0.55 (within) |
NCI60 Strict Holdout (nci60_strict_holdout) | Very Hard | genomics | Affy HG-U95 | 64 / 6,830 | P15_lowpn_all_features | 0.30-0.50 (above) |
NCI9 (60 x 9712, 9-class) (nci9_60_9class) | Hard | genomics | cDNA | 60 / 5,726 | V20_F01_flaml_30s_mnpo_diag24 | 0.50-0.70 (within) |
ORLraws10P (Face) (orlraws10p) | Easy | non_genomic | cDNA | 100 / 10,304 | A01_simple_anchor_after_fs | 0.95-1.00 (within) |
Ovarian Cancer (Petricoin) (ovarian_petricoin) | Easy | genomics | mass-spec | 253 / 15,154 | A01_simple_anchor_after_fs | 0.95-1.00 (within) |
PCMAC Text (Scikit-feature) (rv_pcmac) | Medium | genomics | cDNA | 1,943 / 3,289 | V21_CURRENT_DEFAULT | 0.80-0.90 (within) |
pixraw10P (Face) (pixraw10p) | Easy | non_genomic | cDNA | 100 / 10,000 | A03_ref_anchor_after_fs | 0.96-1.00 (within) |
Prostate Cancer (Singh) (prostate_singh) | Easy | genomics | Affy HG-U95 | 102 / 12,600 | A07_ref_before_fs | 0.88-0.95 (above) |
RELATHE Text (Scikit-feature) (rv_relathe) | Medium | genomics | cDNA | 1,427 / 4,322 | V21_WINNER | 0.82-0.92 (within) |
SMK_CAN_187 (Smoking) (smk_can_187) | Medium | genomics | Affy HG-U133A | 187 / 19,993 | M_SCAFFOLD_subspace_stability | 0.65-0.80 (within) |
SRBCT (Khan) (srbct_khan) | Easy | genomics | cDNA | 83 / 2,308 | A01_simple_anchor_after_fs | 0.92-1.00 (within) |
TCGA-BRCA Breast Cancer (Xena, PAM50 5-subtype) (xena_tcga_brca) | Hard | genomics | cDNA | 956 / 20,530 | A07_ref_before_fs | 0.55-0.72 (above) |
TCGA-COAD Colorectal (Xena, histological 2-class) (xena_tcga_coad_cms) | Hard | genomics | cDNA | 323 / 20,530 | V19_C01_old_regime_legacy_full64 | 0.58-0.75 (above) |
TCGA-GBM Glioblastoma (Xena, 4-subtype) (xena_tcga_gbm) | Hard | genomics | cDNA | 164 / 20,530 | V19_C01_old_regime_legacy_full64 | 0.55-0.72 (above) |
TCGA-HNSC Head & Neck (Xena, HPV binary) (xena_tcga_hnsc_hpv) | Medium | genomics | cDNA | 114 / 20,530 | A05_skip_df_ref | 0.82-0.93 (above) |
TCGA-KIRC Kidney Clear Cell (Xena, stage-based) (xena_tcga_kirc) | Hard | genomics | cDNA | 606 / 20,530 | V21_WINNER | 0.50-0.68 (below) |
TCGA-LGG Lower Grade Glioma (Xena, 3-subtype) (xena_tcga_lgg) | Hard | genomics | cDNA | 529 / 20,530 | V19_C04_new_regime_mnpo_full64 | 0.55-0.72 (within) |
TCGA-LIHC Liver (Xena, histologic grade 3-class) (xena_tcga_lihc) | Hard | genomics | cDNA | 415 / 20,530 | V19_C03_old_regime_mnpo_full64 | 0.52-0.68 (within) |
TCGA-LUAD Lung Adenocarcinoma (Xena, 3-subtype) (xena_tcga_luad) | Hard | genomics | cDNA | 275 / 20,530 | A07_ref_before_fs | 0.55-0.70 (above) |
TCGA-OV Ovarian (Xena, histologic grade G2 vs G3) (xena_tcga_ov) | Hard | genomics | cDNA | 299 / 20,530 | V21_WINNER_PREFILTER | 0.48-0.65 (above) |
TCGA-PRAD Prostate (Xena, Gleason score) (xena_tcga_prad) | Hard | genomics | cDNA | 550 / 20,530 | V19_C04_new_regime_mnpo_full64 | 0.45-0.65 (within) |
TCGA-SKCM Melanoma (Xena, primary vs metastatic) (xena_tcga_skcm) | Hard | genomics | cDNA | 472 / 20,530 | V21_WINNER_METASEL | 0.78-0.90 (within) |
TCGA-STAD Stomach (Xena, histological 3-group) (xena_tcga_stad) | Hard | genomics | cDNA | 448 / 20,530 | V20_L01_lr_prior_reduced_full64 | 0.58-0.75 (within) |
TCGA-UCEC Uterine (Xena, histological 3-type) (xena_tcga_ucec) | Hard | genomics | cDNA | 190 / 20,530 | V21_WINNER | 0.50-0.68 (above) |
TOX_171 (Scikit-feature) (tox_171) | Medium | genomics | cDNA | 171 / 5,748 | A01_simple_anchor_after_fs | 0.84-0.93 (above) |
USPS Handwritten Digits (Scikit-feature) (rv_usps) | Easy | genomics | cDNA | 9,298 / 256 | V21_CURRENT_DEFAULT | 0.90-0.97 (within) |
warpAR10P (Scikit-feature, Face) (rv_warpar10p) | Medium | genomics | cDNA | 130 / 2,400 | V21_CURRENT_DEFAULT | 0.70-0.85 (above) |
warpPIE10P (Face) (warp_pie10p) | Easy | non_genomic | cDNA | 210 / 2,420 | A07_ref_before_fs | 0.95-1.00 (within) |
Yale Faces (Scikit-feature) (rv_yale) | Hard | genomics | cDNA | 165 / 1,024 | V21_WINNER | 0.55-0.72 (within) |
Documentation and webpages on this site are generated from authoritative internal sources using a combination of deterministic rules and generative AI. Errors are possible. Please report issues via GitHub Discussions or email [email protected].