evaluation · needs review

scGPT + residual geometry: gene-regulatory signal prediction

Evaluation procedure

Asymmetric extraction, PCA-64 centered cosine geometry added to scGPT baseline

Metadata review: needs review. Unreported conditions prevent automatic comparisons.

Release 2026-09-16-d74d282221a9 · 1 evaluation · 1 metric row. Different protocols are not a single leaderboard.

Evaluated configurationMetricEvidence origin

Results grouped by the exact reported evaluation
Metric and finding	Coverage and uncertainty	Evidence
scGPT + residual geometry: gene-regulatory signal prediction Model: scGPT + residual geometry · Benchmark: gene-regulatory signal prediction · Dataset: immune tissue Asymmetric extraction, PCA-64 centered cosine geometry added to scGPT baseline Author-reported evaluation · Evaluation metadata: needs review
0.677 AUROC Unit: fraction · Direction: unknown	Uncertainty: not reported in legacy extract Scored: Not reported · Eligible: Not reported	source checkedResidual-stream geometry of single-cell foundation models carries incremental gene-regulatory signal across tissues · Table 4, Immune row, scGPT > +geom AUROC column Source checking is not independent reproduction.

Release 2026-09-16-d74d282221a9 · Record review: needs review

Technical metadata and extraction receipts

Stable ID: evaluation-lit-b4-013

areas: cells-tissues
tasks: gene-regulatory signal prediction
origin: author_reported
protocol: Asymmetric extraction, PCA-64 centered cosine geometry added to scGPT baseline
version: Not reported
comparison: protocol id: Not reported; dataset version: Not reported; split: Not reported; population: Not reported; inputs: Not reported; adaptation: Not reported; metric implementation: Not reported; aggregation: Not reported; budget: Not reported
missing metadata: model version: not_reported_in_legacy_extract; dataset version: not_reported_in_legacy_extract; split: not_reported_in_legacy_extract

Related records