Configuration

AssayFormer, BPMF embeddings, GRPO, no screen description (AssayBench-Loop)

Configuration of AssayFormer, BPMF embeddings, GRPO, no screen description: BPMF embeddings; supervised training then GRPO; screen description removed.

1 evaluation · 10 results

Overview

An explanatory profile has not yet been reviewed for this record.

Consult the linked sources for architecture or protocol details. Missing evidence is not evidence of a missing capability.

Evaluations and results

1 evaluation · 10 results. Different protocols are not a single leaderboard.

Protocol: AssayBench-Loop adaptive hit discovery: 10 rounds of 100 genes on 20 held-out CRISPR screens · Dataset subset: AssayBench-Loop test split (20 genome-wide CRISPR screens published after 2021)

Sorted by Hit enrichment factor (hits found divided by hits expected under random selection, after 10 rounds of 100 genes; hallucinated genes and unfilled slots count in the denominator, valid genes outside the screen library do not; mean over the 20 AssayBench-Loop test screens) (higher is better). The best value in each column is highlighted. Decimals are rounded for display; each value links to the printed value and its source.

Tested configuration	Hit enrichment factor (hits found divided by hits expected under random selection, after 10 rounds of 100 genes; hallucinated genes and unfilled slots count in the denominator, valid genes outside the screen library do not; mean over the 20 AssayBench-Loop test screens)	Normalized cumulative hits AUC (area under the fraction-of-hits against fraction-of-library-queried curve over 10 rounds, divided by an oracle's area over the same effective budget; mean over the 20 AssayBench-Loop test screens)	Proportion (shortfall: share of acquired genes outside the screen library, including hallucinated gene names; mean over the 20 AssayBench-Loop test screens)	Recall (fraction of the screen's hits acquired within 10 rounds of 100 genes (FH); mean over the 20 AssayBench-Loop test screens)	Annotation overlap ratio (mean pairwise Jaccard overlap of GO biological-process annotations within a batch, divided by that of an equal-size random batch from the screen library, averaged over batches; mean over the 20 AssayBench-Loop test screens)	Effective number of groups (EP-B: Hill number of order 1 at batch scope, rarefied to 30 annotated genes per batch, one random group per gene per draw, 400 Monte Carlo draws, averaged within each screen then across the 20 test screens; not comparable with EP-S or EP-D)	Effective number of groups (EP-D: Hill number of order 1 at dataset scope, rarefied to 6,000 annotated genes across the complete test set, one random group per gene per draw, 300 Monte Carlo draws; not comparable with EP-B or EP-S)	Effective number of groups (EP-S: Hill number of order 1 at screen scope, rarefied to 200 annotated genes per screen, one random group per gene per draw, 400 Monte Carlo draws, averaged across the 20 test screens; not comparable with EP-B or EP-D)	Proportion (share of acquired hits that are DepMap common-essential genes (1,827 genes); mean over the 20 AssayBench-Loop test screens)	Vendi ratio (Vendi score of each batch under a cosine kernel on GenePT embeddings, divided by batch size and averaged over batches; mean over the 20 AssayBench-Loop test screens)
AssayFormer, BPMF embeddings, GRPO, no screen description (AssayBench-Loop)Evaluation	4.52	16.2	5.9%	22.4%	2.55	18.7	63.1	45.7	42.5%	57.7

All 10 result rows with coverage, uncertainty and sources

Filter evaluations

Tested model or methodProtocolDatasetEvaluation setupMetricEvidence origin

Applied filters: All linked evaluations

Exact evaluated configurations and original reported results
Tested configuration	Protocol and dataset	Finding	Evidence and details
Configuration: AssayFormer, BPMF embeddings, GRPO, no screen description (AssayBench-Loop)	Protocol: AssayBench-Loop adaptive hit discovery: 10 rounds of 100 genes on 20 held-out CRISPR screens Dataset subset: AssayBench-Loop test split (20 genome-wide CRISPR screens published after 2021)	4.52 Hit enrichment factor (hits found divided by hits expected under random selection, after 10 rounds of 100 genes; hallucinated genes and unfilled slots count in the denominator, valid genes outside the screen library do not; mean over the 20 AssayBench-Loop test screens) unitless · higher Uncertainty: Not reported by the source: Table 1 and Supplementary Table 4 print means without spread. Coverage: 20 test screens scored	Author-reported evaluation · Source checked Methods, coverage and source AssayFormer, BPMF embeddings, GRPO, no screen description on AssayBench-Loop test (Edwards et al. 2026) ucpheno-20261022-protocol-edwards2026-assaybench-loop-test-10x100 Aggregation: Not reported Biology-in-the-loop: Amortized Adaptive Hit Discovery in CRISPR Screens · Supplementary Table 4, row "BPMF (no desc.) + GRPO" (group 'Ablations'), column "EF"
Configuration: AssayFormer, BPMF embeddings, GRPO, no screen description (AssayBench-Loop)	Protocol: AssayBench-Loop adaptive hit discovery: 10 rounds of 100 genes on 20 held-out CRISPR screens Dataset subset: AssayBench-Loop test split (20 genome-wide CRISPR screens published after 2021)	18.7 Effective number of groups (EP-B: Hill number of order 1 at batch scope, rarefied to 30 annotated genes per batch, one random group per gene per draw, 400 Monte Carlo draws, averaged within each screen then across the 20 test screens; not comparable with EP-S or EP-D) unitless · unknown Uncertainty: Not reported by the source: Table 1 and Supplementary Table 4 print means without spread. Coverage: 20 test screens scored (Appendix H excludes batches or screens with fewer annotated genes than the rarefaction size; how many were excluded is not reported.)	Author-reported evaluation · Source checked Methods, coverage and source AssayFormer, BPMF embeddings, GRPO, no screen description on AssayBench-Loop test (Edwards et al. 2026) ucpheno-20261022-protocol-edwards2026-assaybench-loop-test-10x100 Aggregation: Not reported Biology-in-the-loop: Amortized Adaptive Hit Discovery in CRISPR Screens · Supplementary Table 4, row "BPMF (no desc.) + GRPO" (group 'Ablations'), column "EP-B"
Configuration: AssayFormer, BPMF embeddings, GRPO, no screen description (AssayBench-Loop)	Protocol: AssayBench-Loop adaptive hit discovery: 10 rounds of 100 genes on 20 held-out CRISPR screens Dataset subset: AssayBench-Loop test split (20 genome-wide CRISPR screens published after 2021)	63.1 Effective number of groups (EP-D: Hill number of order 1 at dataset scope, rarefied to 6,000 annotated genes across the complete test set, one random group per gene per draw, 300 Monte Carlo draws; not comparable with EP-B or EP-S) unitless · unknown Uncertainty: Not reported by the source: Table 1 and Supplementary Table 4 print means without spread. Coverage: 20 test screens scored (One value pooled over the 20 test screens (Appendix H), not a per-screen mean as the Table 1 caption says of all metrics. Appendix H also excludes sample sets with fewer annotated genes than the rarefaction size; how many is not reported.)	Author-reported evaluation · Source checked Methods, coverage and source AssayFormer, BPMF embeddings, GRPO, no screen description on AssayBench-Loop test (Edwards et al. 2026) ucpheno-20261022-protocol-edwards2026-assaybench-loop-test-10x100 Aggregation: Not reported Biology-in-the-loop: Amortized Adaptive Hit Discovery in CRISPR Screens · Supplementary Table 4, row "BPMF (no desc.) + GRPO" (group 'Ablations'), column "EP-D"
Configuration: AssayFormer, BPMF embeddings, GRPO, no screen description (AssayBench-Loop)	Protocol: AssayBench-Loop adaptive hit discovery: 10 rounds of 100 genes on 20 held-out CRISPR screens Dataset subset: AssayBench-Loop test split (20 genome-wide CRISPR screens published after 2021)	45.7 Effective number of groups (EP-S: Hill number of order 1 at screen scope, rarefied to 200 annotated genes per screen, one random group per gene per draw, 400 Monte Carlo draws, averaged across the 20 test screens; not comparable with EP-B or EP-D) unitless · unknown Uncertainty: Not reported by the source: Table 1 and Supplementary Table 4 print means without spread. Coverage: 20 test screens scored (Appendix H excludes batches or screens with fewer annotated genes than the rarefaction size; how many were excluded is not reported.)	Author-reported evaluation · Source checked Methods, coverage and source AssayFormer, BPMF embeddings, GRPO, no screen description on AssayBench-Loop test (Edwards et al. 2026) ucpheno-20261022-protocol-edwards2026-assaybench-loop-test-10x100 Aggregation: Not reported Biology-in-the-loop: Amortized Adaptive Hit Discovery in CRISPR Screens · Supplementary Table 4, row "BPMF (no desc.) + GRPO" (group 'Ablations'), column "EP-S"
Configuration: AssayFormer, BPMF embeddings, GRPO, no screen description (AssayBench-Loop)	Protocol: AssayBench-Loop adaptive hit discovery: 10 rounds of 100 genes on 20 held-out CRISPR screens Dataset subset: AssayBench-Loop test split (20 genome-wide CRISPR screens published after 2021)	42.5% Proportion (share of acquired hits that are DepMap common-essential genes (1,827 genes); mean over the 20 AssayBench-Loop test screens) percent · unknown Uncertainty: Not reported by the source: Table 1 and Supplementary Table 4 print means without spread. Coverage: 20 test screens scored	Author-reported evaluation · Source checked Methods, coverage and source AssayFormer, BPMF embeddings, GRPO, no screen description on AssayBench-Loop test (Edwards et al. 2026) ucpheno-20261022-protocol-edwards2026-assaybench-loop-test-10x100 Aggregation: Not reported Biology-in-the-loop: Amortized Adaptive Hit Discovery in CRISPR Screens · Supplementary Table 4, row "BPMF (no desc.) + GRPO" (group 'Ablations'), column "Ess. (%)"
Configuration: AssayFormer, BPMF embeddings, GRPO, no screen description (AssayBench-Loop)	Protocol: AssayBench-Loop adaptive hit discovery: 10 rounds of 100 genes on 20 held-out CRISPR screens Dataset subset: AssayBench-Loop test split (20 genome-wide CRISPR screens published after 2021)	22.4% Recall (fraction of the screen's hits acquired within 10 rounds of 100 genes (FH); mean over the 20 AssayBench-Loop test screens) percent · higher Uncertainty: Not reported by the source: Table 1 and Supplementary Table 4 print means without spread. Coverage: 20 test screens scored	Author-reported evaluation · Source checked Methods, coverage and source AssayFormer, BPMF embeddings, GRPO, no screen description on AssayBench-Loop test (Edwards et al. 2026) ucpheno-20261022-protocol-edwards2026-assaybench-loop-test-10x100 Aggregation: Not reported Biology-in-the-loop: Amortized Adaptive Hit Discovery in CRISPR Screens · Supplementary Table 4, row "BPMF (no desc.) + GRPO" (group 'Ablations'), column "Frac. hits (%)"
Configuration: AssayFormer, BPMF embeddings, GRPO, no screen description (AssayBench-Loop)	Protocol: AssayBench-Loop adaptive hit discovery: 10 rounds of 100 genes on 20 held-out CRISPR screens Dataset subset: AssayBench-Loop test split (20 genome-wide CRISPR screens published after 2021)	16.2 Normalized cumulative hits AUC (area under the fraction-of-hits against fraction-of-library-queried curve over 10 rounds, divided by an oracle's area over the same effective budget; mean over the 20 AssayBench-Loop test screens) unitless · higher Uncertainty: Not reported by the source: Table 1 and Supplementary Table 4 print means without spread. Coverage: 20 test screens scored	Author-reported evaluation · Source checked Methods, coverage and source AssayFormer, BPMF embeddings, GRPO, no screen description on AssayBench-Loop test (Edwards et al. 2026) ucpheno-20261022-protocol-edwards2026-assaybench-loop-test-10x100 Aggregation: Not reported Biology-in-the-loop: Amortized Adaptive Hit Discovery in CRISPR Screens · Supplementary Table 4, row "BPMF (no desc.) + GRPO" (group 'Ablations'), column "nAUC (%)"
Configuration: AssayFormer, BPMF embeddings, GRPO, no screen description (AssayBench-Loop)	Protocol: AssayBench-Loop adaptive hit discovery: 10 rounds of 100 genes on 20 held-out CRISPR screens Dataset subset: AssayBench-Loop test split (20 genome-wide CRISPR screens published after 2021)	2.55 Annotation overlap ratio (mean pairwise Jaccard overlap of GO biological-process annotations within a batch, divided by that of an equal-size random batch from the screen library, averaged over batches; mean over the 20 AssayBench-Loop test screens) unitless · unknown Uncertainty: Not reported by the source: Table 1 and Supplementary Table 4 print means without spread. Coverage: 20 test screens scored	Author-reported evaluation · Source checked Methods, coverage and source AssayFormer, BPMF embeddings, GRPO, no screen description on AssayBench-Loop test (Edwards et al. 2026) ucpheno-20261022-protocol-edwards2026-assaybench-loop-test-10x100 Aggregation: Not reported Biology-in-the-loop: Amortized Adaptive Hit Discovery in CRISPR Screens · Supplementary Table 4, row "BPMF (no desc.) + GRPO" (group 'Ablations'), column "Path. Ov."
Configuration: AssayFormer, BPMF embeddings, GRPO, no screen description (AssayBench-Loop)	Protocol: AssayBench-Loop adaptive hit discovery: 10 rounds of 100 genes on 20 held-out CRISPR screens Dataset subset: AssayBench-Loop test split (20 genome-wide CRISPR screens published after 2021)	5.9% Proportion (shortfall: share of acquired genes outside the screen library, including hallucinated gene names; mean over the 20 AssayBench-Loop test screens) percent · lower Uncertainty: Not reported by the source: Table 1 and Supplementary Table 4 print means without spread. Coverage: 20 test screens scored	Author-reported evaluation · Source checked Methods, coverage and source AssayFormer, BPMF embeddings, GRPO, no screen description on AssayBench-Loop test (Edwards et al. 2026) ucpheno-20261022-protocol-edwards2026-assaybench-loop-test-10x100 Aggregation: Not reported Biology-in-the-loop: Amortized Adaptive Hit Discovery in CRISPR Screens · Supplementary Table 4, row "BPMF (no desc.) + GRPO" (group 'Ablations'), column "Shortfall (%)"
Configuration: AssayFormer, BPMF embeddings, GRPO, no screen description (AssayBench-Loop)	Protocol: AssayBench-Loop adaptive hit discovery: 10 rounds of 100 genes on 20 held-out CRISPR screens Dataset subset: AssayBench-Loop test split (20 genome-wide CRISPR screens published after 2021)	57.7 Vendi ratio (Vendi score of each batch under a cosine kernel on GenePT embeddings, divided by batch size and averaged over batches; mean over the 20 AssayBench-Loop test screens) unitless · unknown Uncertainty: Not reported by the source: Table 1 and Supplementary Table 4 print means without spread. Coverage: 20 test screens scored	Author-reported evaluation · Source checked Methods, coverage and source AssayFormer, BPMF embeddings, GRPO, no screen description on AssayBench-Loop test (Edwards et al. 2026) ucpheno-20261022-protocol-edwards2026-assaybench-loop-test-10x100 Aggregation: Not reported Biology-in-the-loop: Amortized Adaptive Hit Discovery in CRISPR Screens · Supplementary Table 4, row "BPMF (no desc.) + GRPO" (group 'Ablations'), column "Vendi (%)"

Source checking is not independent reproduction. Release 2026-10-23-5347695500ed.

Use this configuration

How it works, versions and access

Additional methodological details, versions and access terms have not yet been extracted into a structured profile for this configuration. Consult the linked sources and evaluation records below.

Evidence

Source checking verifies the cited claim or transcription. It does not establish independent reproduction.

Evidence table

Inspect claims, sources and review details

Trace each statement to its source and review. A context-only reference supports the record generally; it does not verify an individual field. Source checking does not reproduce an experiment.

One row per statement and cited source. Multiple citations are not independent evaluations. Shared locators are labelled explicitly.

0 evidence rows matching the loaded filters

Claims, original sources and review scope · Release 2026-10-23-5347695500ed
Property and statement	Original source and location	Review and provenance

No evidence rows match these filters. Choose another scope or clear the search.

How to read the evidence

Sources and history

Release 2026-10-23-5347695500ed · Record review: source checked

1 source record and release history

Biology-in-the-loop: Amortized Adaptive Hit Discovery in CRISPR Screens · Original source · arXiv:2609.11877 version 1, submitted 2026-09-10; not peer reviewed

Download this release (gzip)

Technical metadata and extraction receipts

Stable ID: ucpheno-20261022-config-edwards2026-assayformer-no-desc-grpo

areas: cells-tissues
contexts: research
method types: specialist
reported name: BPMF (no desc.) + GRPO
foundation model eligible: false
parameters: BPMF embeddings; supervised training then GRPO; screen description removed
source locator: Table 1 and Supplementary Table 4 row labels; Methods 4.6-4.9
missing metadata: access reuse: reason: unextracted; note: Page 1 shows Website, GitHub, PyPI and Models & Data link labels; their URLs were not resolved from the text layer; version: reason: unreported; note: No release, commit or checkpoint is printed for this configuration
provenance note: Configuration as run in the AssayLoop comparison (arXiv:2609.11877v1).

Related records

configuration of: AssayFormer
system: AssayFormer, BPMF embeddings, GRPO, no screen description on AssayBench-Loop test (Edwards et al. 2026)

Suggest a correction