MassSpecGym formula-known retrieval (bonus), Hit rate@1 (MassSpecGym in the Wild Table 3); MassSpecGym as audited in arXiv:2606.19624v1

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.

18 recorded evaluations, 18 metric rows. A comparison chart has not yet been validated for these results. The table retains the individual findings and their sources.

View coverage and remaining gaps across all benchmarks

Results

Results are available, but no reviewed comparison panel is linked in this release.

All evaluations

18 evaluations · 18 results. Different protocols are not a single leaderboard.

Protocol: MassSpecGym formula-known retrieval (bonus), Hit rate@1 (MassSpecGym in the Wild Table 3) · Dataset: MassSpecGym test fold, formula-known retrieval candidates (bonus) (MassSpecGym in the Wild)

Sorted by Top-1 accuracy (retrieval) (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	Top-1 accuracy (retrieval)
ChemBERTa SMILES classifier (spectrum-free), original candidate canonicalization (MassSpecGym in the Wild)Evaluation	99.66 (99.50-99.79)% (best in column)
DreaMS spectral encoder aligned to frozen ChemBERTa, original candidate canonicalization (MassSpecGym in the Wild)Evaluation	82.41 (81.46-83.33)%
ChemFormer two-stage retrieval, original candidate canonicalization (MassSpecGym in the Wild)Evaluation	63.89 (62.71-65.03)%
MIST, FRIGID checkpoint (vFRIGID) (MassSpecGym in the Wild)Evaluation	53.76 (52.51-55.02)%
PubChem default ranking (spectrum-blind) (MassSpecGym in the Wild)Evaluation	49.98 (48.72-51.20)%
MIST, DiffMS checkpoint (vDiffMS) (MassSpecGym in the Wild)Evaluation	46.42 (45.07-47.58)%
FRIGID generative retrieval (MassSpecGym in the Wild)Evaluation	45.27 (44.06-46.43)%
MIST + MolForge generative retrieval, reimplemented by the authors (MassSpecGym in the Wild)Evaluation	39.09 (37.86-40.26)%
ICEBERG forward-model retrieval with imputed collision energies (MassSpecGym in the Wild)Evaluation	36.18 (34.96–37.32)%
DiffMS generative retrieval (MassSpecGym in the Wild)Evaluation	32.29 (31.13-33.45)%
FLARE (MassSpecGym in the Wild)Evaluation	22.66 (21.63–23.74)%
MVP (MassSpecGym in the Wild)Evaluation	13.96 (13.10-14.82)%
JESTR (MassSpecGym in the Wild)Evaluation	11.82 (11.03-12.68)%
Nearest neighbour, reimplemented by the authors (MassSpecGym in the Wild)Evaluation	9.58 (8.84-10.30)%
ChemBERTa SMILES classifier (spectrum-free), corrected candidate canonicalization (MassSpecGym in the Wild)Evaluation	8.2 (7.53-8.89)%
ChemFormer two-stage retrieval, corrected candidate canonicalization (MassSpecGym in the Wild)Evaluation	7.60 (6.94-8.31)%
DreaMS spectral encoder aligned to frozen ChemBERTa, corrected candidate canonicalization (MassSpecGym in the Wild)Evaluation	6.70 (6.09-7.35)%
Ranking by number of chiral atoms (spectrum-blind) (MassSpecGym in the Wild)Evaluation	4.88 (4.36-5.44)%

All 18 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: ChemBERTa SMILES classifier (spectrum-free), corrected candidate canonicalization (MassSpecGym in the Wild)	Protocol: MassSpecGym formula-known retrieval (bonus), Hit rate@1 (MassSpecGym in the Wild Table 3) Dataset: MassSpecGym test fold, formula-known retrieval candidates (bonus) (MassSpecGym in the Wild)	8.2 (7.53-8.89)% Top-1 accuracy (retrieval) percent · higher Uncertainty: 99.9% CI 7.53 to 8.89. Range printed in parentheses after the mean. Appendix 'Confidence Intervals': 99.9% bootstrap interval over 20,000 resamples of the test set. Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ChemBERTa SMILES classifier (spectrum-free), corrected candidate canonicalization on MassSpecGym formula-retrieval (Hit rate@1) lit-otheromics-20261011-msgwild-protocol-formula-retrieval-hit-rate-1 Aggregation: Not reported MassSpecGym in the Wild: Uncovering and Correcting Evaluation Pitfalls in AI-Driven Molecule Discovery · Table 3, section 'Exploiting Canonicalization Artifacts (Not Valid Under Corrected Candidate Canonicalization)', row '(3) ChemBERTa SMILES classifier (corrected)', column 'Hit rate@1'
Configuration: ChemBERTa SMILES classifier (spectrum-free), original candidate canonicalization (MassSpecGym in the Wild)	Protocol: MassSpecGym formula-known retrieval (bonus), Hit rate@1 (MassSpecGym in the Wild Table 3) Dataset: MassSpecGym test fold, formula-known retrieval candidates (bonus) (MassSpecGym in the Wild)	99.66 (99.50-99.79)% Top-1 accuracy (retrieval) percent · higher Uncertainty: 99.9% CI 99.50 to 99.79. Range printed in parentheses after the mean. Appendix 'Confidence Intervals': 99.9% bootstrap interval over 20,000 resamples of the test set. Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ChemBERTa SMILES classifier (spectrum-free), original candidate canonicalization on MassSpecGym formula-retrieval (Hit rate@1) lit-otheromics-20261011-msgwild-protocol-formula-retrieval-hit-rate-1 Aggregation: Not reported MassSpecGym in the Wild: Uncovering and Correcting Evaluation Pitfalls in AI-Driven Molecule Discovery · Table 3, section 'Exploiting Canonicalization Artifacts (Not Valid Under Corrected Candidate Canonicalization)', row '(3) ChemBERTa SMILES classifier (inflated)', column 'Hit rate@1'
Configuration: ChemFormer two-stage retrieval, corrected candidate canonicalization (MassSpecGym in the Wild)	Protocol: MassSpecGym formula-known retrieval (bonus), Hit rate@1 (MassSpecGym in the Wild Table 3) Dataset: MassSpecGym test fold, formula-known retrieval candidates (bonus) (MassSpecGym in the Wild)	7.60 (6.94-8.31)% Top-1 accuracy (retrieval) percent · higher Uncertainty: 99.9% CI 6.94 to 8.31. Range printed in parentheses after the mean. Appendix 'Confidence Intervals': 99.9% bootstrap interval over 20,000 resamples of the test set. Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ChemFormer two-stage retrieval, corrected candidate canonicalization on MassSpecGym formula-retrieval (Hit rate@1) lit-otheromics-20261011-msgwild-protocol-formula-retrieval-hit-rate-1 Aggregation: Not reported MassSpecGym in the Wild: Uncovering and Correcting Evaluation Pitfalls in AI-Driven Molecule Discovery · Table 3, section 'Exploiting Canonicalization Artifacts (Not Valid Under Corrected Candidate Canonicalization)', row '(2) ChemFormer Two-Stage Retrieval (corrected)', column 'Hit rate@1'
Configuration: ChemFormer two-stage retrieval, original candidate canonicalization (MassSpecGym in the Wild)	Protocol: MassSpecGym formula-known retrieval (bonus), Hit rate@1 (MassSpecGym in the Wild Table 3) Dataset: MassSpecGym test fold, formula-known retrieval candidates (bonus) (MassSpecGym in the Wild)	63.89 (62.71-65.03)% Top-1 accuracy (retrieval) percent · higher Uncertainty: 99.9% CI 62.71 to 65.03. Range printed in parentheses after the mean. Appendix 'Confidence Intervals': 99.9% bootstrap interval over 20,000 resamples of the test set. Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ChemFormer two-stage retrieval, original candidate canonicalization on MassSpecGym formula-retrieval (Hit rate@1) lit-otheromics-20261011-msgwild-protocol-formula-retrieval-hit-rate-1 Aggregation: Not reported MassSpecGym in the Wild: Uncovering and Correcting Evaluation Pitfalls in AI-Driven Molecule Discovery · Table 3, section 'Exploiting Canonicalization Artifacts (Not Valid Under Corrected Candidate Canonicalization)', row '(2) ChemFormer Two-Stage Retrieval (inflated)', column 'Hit rate@1'
Configuration: DiffMS generative retrieval (MassSpecGym in the Wild)	Protocol: MassSpecGym formula-known retrieval (bonus), Hit rate@1 (MassSpecGym in the Wild Table 3) Dataset: MassSpecGym test fold, formula-known retrieval candidates (bonus) (MassSpecGym in the Wild)	32.29 (31.13-33.45)% Top-1 accuracy (retrieval) percent · higher Uncertainty: 99.9% CI 31.13 to 33.45. Range printed in parentheses after the mean. Table 3 caption: 99.9% BCa bootstrap intervals over 20,000 resamples, stated for all bracketed values. Value repeated from the cited work; that work's own interval settings were not checked. Coverage: Not reported scored / Not reported eligible	Result quoted from another source · Source checked Methods, coverage and source DiffMS generative retrieval on MassSpecGym formula-retrieval (Hit rate@1) lit-otheromics-20261011-msgwild-protocol-formula-retrieval-hit-rate-1 Aggregation: Not reported MassSpecGym in the Wild: Uncovering and Correcting Evaluation Pitfalls in AI-Driven Molecule Discovery · Table 3, section 'Reported Performance in Retrieval Task w/ Bonus (i.e., Formula Known)', row 'DiffMS (generative retrieval)', column 'Hit rate@1'
Configuration: DreaMS spectral encoder aligned to frozen ChemBERTa, corrected candidate canonicalization (MassSpecGym in the Wild)	Protocol: MassSpecGym formula-known retrieval (bonus), Hit rate@1 (MassSpecGym in the Wild Table 3) Dataset: MassSpecGym test fold, formula-known retrieval candidates (bonus) (MassSpecGym in the Wild)	6.70 (6.09-7.35)% Top-1 accuracy (retrieval) percent · higher Uncertainty: 99.9% CI 6.09 to 7.35. Range printed in parentheses after the mean. Appendix 'Confidence Intervals': 99.9% bootstrap interval over 20,000 resamples of the test set. Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source DreaMS spectral encoder aligned to frozen ChemBERTa, corrected candidate canonicalization on MassSpecGym formula-retrieval (Hit rate@1) lit-otheromics-20261011-msgwild-protocol-formula-retrieval-hit-rate-1 Aggregation: Not reported MassSpecGym in the Wild: Uncovering and Correcting Evaluation Pitfalls in AI-Driven Molecule Discovery · Table 3, section 'Exploiting Canonicalization Artifacts (Not Valid Under Corrected Candidate Canonicalization)', row '(1) DreaMS + ChemBERTa Alignment (corrected)', column 'Hit rate@1'
Configuration: DreaMS spectral encoder aligned to frozen ChemBERTa, original candidate canonicalization (MassSpecGym in the Wild)	Protocol: MassSpecGym formula-known retrieval (bonus), Hit rate@1 (MassSpecGym in the Wild Table 3) Dataset: MassSpecGym test fold, formula-known retrieval candidates (bonus) (MassSpecGym in the Wild)	82.41 (81.46-83.33)% Top-1 accuracy (retrieval) percent · higher Uncertainty: 99.9% CI 81.46 to 83.33. Range printed in parentheses after the mean. Appendix 'Confidence Intervals': 99.9% bootstrap interval over 20,000 resamples of the test set. Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source DreaMS spectral encoder aligned to frozen ChemBERTa, original candidate canonicalization on MassSpecGym formula-retrieval (Hit rate@1) lit-otheromics-20261011-msgwild-protocol-formula-retrieval-hit-rate-1 Aggregation: Not reported MassSpecGym in the Wild: Uncovering and Correcting Evaluation Pitfalls in AI-Driven Molecule Discovery · Table 3, section 'Exploiting Canonicalization Artifacts (Not Valid Under Corrected Candidate Canonicalization)', row '(1) DreaMS + ChemBERTa Alignment (inflated)', column 'Hit rate@1'
Configuration: FLARE (MassSpecGym in the Wild)	Protocol: MassSpecGym formula-known retrieval (bonus), Hit rate@1 (MassSpecGym in the Wild Table 3) Dataset: MassSpecGym test fold, formula-known retrieval candidates (bonus) (MassSpecGym in the Wild)	22.66 (21.63–23.74)% Top-1 accuracy (retrieval) percent · higher Uncertainty: 99.9% CI 21.63 to 23.74. Range printed in parentheses after the mean. Table 3 caption: 99.9% BCa bootstrap intervals over 20,000 resamples, stated for all bracketed values. Value repeated from the cited work; that work's own interval settings were not checked. Coverage: Not reported scored / Not reported eligible	Result quoted from another source · Source checked Methods, coverage and source FLARE on MassSpecGym formula-retrieval (Hit rate@1) lit-otheromics-20261011-msgwild-protocol-formula-retrieval-hit-rate-1 Aggregation: Not reported MassSpecGym in the Wild: Uncovering and Correcting Evaluation Pitfalls in AI-Driven Molecule Discovery · Table 3, section 'Reported Performance in Retrieval Task w/ Bonus (i.e., Formula Known)', row 'FLARE', column 'Hit rate@1'
Configuration: FRIGID generative retrieval (MassSpecGym in the Wild)	Protocol: MassSpecGym formula-known retrieval (bonus), Hit rate@1 (MassSpecGym in the Wild Table 3) Dataset: MassSpecGym test fold, formula-known retrieval candidates (bonus) (MassSpecGym in the Wild)	45.27 (44.06-46.43)% Top-1 accuracy (retrieval) percent · higher Uncertainty: 99.9% CI 44.06 to 46.43. Range printed in parentheses after the mean. Table 3 caption: 99.9% BCa bootstrap intervals over 20,000 resamples, stated for all bracketed values. Value repeated from the cited work; that work's own interval settings were not checked. Coverage: Not reported scored / Not reported eligible	Result quoted from another source · Source checked Methods, coverage and source FRIGID generative retrieval on MassSpecGym formula-retrieval (Hit rate@1) lit-otheromics-20261011-msgwild-protocol-formula-retrieval-hit-rate-1 Aggregation: Not reported MassSpecGym in the Wild: Uncovering and Correcting Evaluation Pitfalls in AI-Driven Molecule Discovery · Table 3, section 'Reported Performance in Retrieval Task w/ Bonus (i.e., Formula Known)', row 'FRIGID (generative retrieval)', column 'Hit rate@1'
Configuration: ICEBERG forward-model retrieval with imputed collision energies (MassSpecGym in the Wild)	Protocol: MassSpecGym formula-known retrieval (bonus), Hit rate@1 (MassSpecGym in the Wild Table 3) Dataset: MassSpecGym test fold, formula-known retrieval candidates (bonus) (MassSpecGym in the Wild)	36.18 (34.96–37.32)% Top-1 accuracy (retrieval) percent · higher Uncertainty: 99.9% CI 34.96 to 37.32. Range printed in parentheses after the mean. Appendix 'Confidence Intervals': 99.9% bootstrap interval over 20,000 resamples of the test set. Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ICEBERG forward-model retrieval with imputed collision energies on MassSpecGym formula-retrieval (Hit rate@1) lit-otheromics-20261011-msgwild-protocol-formula-retrieval-hit-rate-1 Aggregation: Not reported MassSpecGym in the Wild: Uncovering and Correcting Evaluation Pitfalls in AI-Driven Molecule Discovery · Table 3, section 'Reported Performance in Retrieval Task w/ Bonus (i.e., Formula Known)', row 'ICEBERG (forward model retrieval)', column 'Hit rate@1'
Configuration: JESTR (MassSpecGym in the Wild)	Protocol: MassSpecGym formula-known retrieval (bonus), Hit rate@1 (MassSpecGym in the Wild Table 3) Dataset: MassSpecGym test fold, formula-known retrieval candidates (bonus) (MassSpecGym in the Wild)	11.82 (11.03-12.68)% Top-1 accuracy (retrieval) percent · higher Uncertainty: 99.9% CI 11.03 to 12.68. Range printed in parentheses after the mean. Table 3 caption: 99.9% BCa bootstrap intervals over 20,000 resamples, stated for all bracketed values. Value repeated from the cited work; that work's own interval settings were not checked. Coverage: Not reported scored / Not reported eligible	Result quoted from another source · Source checked Methods, coverage and source JESTR on MassSpecGym formula-retrieval (Hit rate@1) lit-otheromics-20261011-msgwild-protocol-formula-retrieval-hit-rate-1 Aggregation: Not reported MassSpecGym in the Wild: Uncovering and Correcting Evaluation Pitfalls in AI-Driven Molecule Discovery · Table 3, section 'Reported Performance in Retrieval Task w/ Bonus (i.e., Formula Known)', row 'JESTR', column 'Hit rate@1'
Configuration: MIST, DiffMS checkpoint (vDiffMS) (MassSpecGym in the Wild)	Protocol: MassSpecGym formula-known retrieval (bonus), Hit rate@1 (MassSpecGym in the Wild Table 3) Dataset: MassSpecGym test fold, formula-known retrieval candidates (bonus) (MassSpecGym in the Wild)	46.42 (45.07-47.58)% Top-1 accuracy (retrieval) percent · higher Uncertainty: 99.9% CI 45.07 to 47.58. Range printed in parentheses after the mean. Table 3 caption: 99.9% BCa bootstrap intervals over 20,000 resamples, stated for all bracketed values. Value repeated from the cited work; that work's own interval settings were not checked. Coverage: Not reported scored / Not reported eligible	Result quoted from another source · Source checked Methods, coverage and source MIST, DiffMS checkpoint (vDiffMS) on MassSpecGym formula-retrieval (Hit rate@1) lit-otheromics-20261011-msgwild-protocol-formula-retrieval-hit-rate-1 Aggregation: Not reported MassSpecGym in the Wild: Uncovering and Correcting Evaluation Pitfalls in AI-Driven Molecule Discovery · Table 3, section 'Reported Performance in Retrieval Task w/ Bonus (i.e., Formula Known)', row 'MIST vDiffMS', column 'Hit rate@1'
Configuration: MIST, FRIGID checkpoint (vFRIGID) (MassSpecGym in the Wild)	Protocol: MassSpecGym formula-known retrieval (bonus), Hit rate@1 (MassSpecGym in the Wild Table 3) Dataset: MassSpecGym test fold, formula-known retrieval candidates (bonus) (MassSpecGym in the Wild)	53.76 (52.51-55.02)% Top-1 accuracy (retrieval) percent · higher Uncertainty: 99.9% CI 52.51 to 55.02. Range printed in parentheses after the mean. Table 3 caption: 99.9% BCa bootstrap intervals over 20,000 resamples, stated for all bracketed values. Value repeated from the cited work; that work's own interval settings were not checked. Coverage: Not reported scored / Not reported eligible	Result quoted from another source · Source checked Methods, coverage and source MIST, FRIGID checkpoint (vFRIGID) on MassSpecGym formula-retrieval (Hit rate@1) lit-otheromics-20261011-msgwild-protocol-formula-retrieval-hit-rate-1 Aggregation: Not reported MassSpecGym in the Wild: Uncovering and Correcting Evaluation Pitfalls in AI-Driven Molecule Discovery · Table 3, section 'Reported Performance in Retrieval Task w/ Bonus (i.e., Formula Known)', row 'MIST vFRIGID', column 'Hit rate@1'
Configuration: MIST + MolForge generative retrieval, reimplemented by the authors (MassSpecGym in the Wild)	Protocol: MassSpecGym formula-known retrieval (bonus), Hit rate@1 (MassSpecGym in the Wild Table 3) Dataset: MassSpecGym test fold, formula-known retrieval candidates (bonus) (MassSpecGym in the Wild)	39.09 (37.86-40.26)% Top-1 accuracy (retrieval) percent · higher Uncertainty: 99.9% CI 37.86 to 40.26. Range printed in parentheses after the mean. Appendix 'Confidence Intervals': 99.9% bootstrap interval over 20,000 resamples of the test set. Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source MIST + MolForge generative retrieval, reimplemented by the authors on MassSpecGym formula-retrieval (Hit rate@1) lit-otheromics-20261011-msgwild-protocol-formula-retrieval-hit-rate-1 Aggregation: Not reported MassSpecGym in the Wild: Uncovering and Correcting Evaluation Pitfalls in AI-Driven Molecule Discovery · Table 3, section 'Reported Performance in Retrieval Task w/ Bonus (i.e., Formula Known)', row 'MIST+MolForge (generative retrieval)†', column 'Hit rate@1'
Configuration: MVP (MassSpecGym in the Wild)	Protocol: MassSpecGym formula-known retrieval (bonus), Hit rate@1 (MassSpecGym in the Wild Table 3) Dataset: MassSpecGym test fold, formula-known retrieval candidates (bonus) (MassSpecGym in the Wild)	13.96 (13.10-14.82)% Top-1 accuracy (retrieval) percent · higher Uncertainty: 99.9% CI 13.10 to 14.82. Range printed in parentheses after the mean. Table 3 caption: 99.9% BCa bootstrap intervals over 20,000 resamples, stated for all bracketed values. Value repeated from the cited work; that work's own interval settings were not checked. Coverage: Not reported scored / Not reported eligible	Result quoted from another source · Source checked Methods, coverage and source MVP on MassSpecGym formula-retrieval (Hit rate@1) lit-otheromics-20261011-msgwild-protocol-formula-retrieval-hit-rate-1 Aggregation: Not reported MassSpecGym in the Wild: Uncovering and Correcting Evaluation Pitfalls in AI-Driven Molecule Discovery · Table 3, section 'Reported Performance in Retrieval Task w/ Bonus (i.e., Formula Known)', row 'MVP', column 'Hit rate@1'
Configuration: Nearest neighbour, reimplemented by the authors (MassSpecGym in the Wild)	Protocol: MassSpecGym formula-known retrieval (bonus), Hit rate@1 (MassSpecGym in the Wild Table 3) Dataset: MassSpecGym test fold, formula-known retrieval candidates (bonus) (MassSpecGym in the Wild)	9.58 (8.84-10.30)% Top-1 accuracy (retrieval) percent · higher Uncertainty: 99.9% CI 8.84 to 10.30. Range printed in parentheses after the mean. Appendix 'Confidence Intervals': 99.9% bootstrap interval over 20,000 resamples of the test set. Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source Nearest neighbour, reimplemented by the authors on MassSpecGym formula-retrieval (Hit rate@1) lit-otheromics-20261011-msgwild-protocol-formula-retrieval-hit-rate-1 Aggregation: Not reported MassSpecGym in the Wild: Uncovering and Correcting Evaluation Pitfalls in AI-Driven Molecule Discovery · Table 3, section 'Reported Performance in Retrieval Task w/ Bonus (i.e., Formula Known)', row 'Nearest Neighbor†', column 'Hit rate@1'
Configuration: PubChem default ranking (spectrum-blind) (MassSpecGym in the Wild)	Protocol: MassSpecGym formula-known retrieval (bonus), Hit rate@1 (MassSpecGym in the Wild Table 3) Dataset: MassSpecGym test fold, formula-known retrieval candidates (bonus) (MassSpecGym in the Wild)	49.98 (48.72-51.20)% Top-1 accuracy (retrieval) percent · higher Uncertainty: 99.9% CI 48.72 to 51.20. Range printed in parentheses after the mean. Appendix 'Confidence Intervals': 99.9% bootstrap interval over 20,000 resamples of the test set. Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source PubChem default ranking (spectrum-blind) on MassSpecGym formula-retrieval (Hit rate@1) lit-otheromics-20261011-msgwild-protocol-formula-retrieval-hit-rate-1 Aggregation: Not reported MassSpecGym in the Wild: Uncovering and Correcting Evaluation Pitfalls in AI-Driven Molecule Discovery · Table 3, section 'Exploiting Other Structure Biases (Spectrum-Blind Stress Test)', row '(4) PubChem default ranking', column 'Hit rate@1'
Configuration: Ranking by number of chiral atoms (spectrum-blind) (MassSpecGym in the Wild)	Protocol: MassSpecGym formula-known retrieval (bonus), Hit rate@1 (MassSpecGym in the Wild Table 3) Dataset: MassSpecGym test fold, formula-known retrieval candidates (bonus) (MassSpecGym in the Wild)	4.88 (4.36-5.44)% Top-1 accuracy (retrieval) percent · higher Uncertainty: 99.9% CI 4.36 to 5.44. Range printed in parentheses after the mean. Appendix 'Confidence Intervals': 99.9% bootstrap interval over 20,000 resamples of the test set. Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source Ranking by number of chiral atoms (spectrum-blind) on MassSpecGym formula-retrieval (Hit rate@1) lit-otheromics-20261011-msgwild-protocol-formula-retrieval-hit-rate-1 Aggregation: Not reported MassSpecGym in the Wild: Uncovering and Correcting Evaluation Pitfalls in AI-Driven Molecule Discovery · Table 3, section 'Exploiting Other Structure Biases (Spectrum-Blind Stress Test)', row '(5) Ranking by num. of chiral atoms', column 'Hit rate@1'

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

Methods and evaluation design

Procedure, tasks and evaluated configurations

Recorded evaluations

Each evaluation records what was tested and under which conditions.

Explore all linked results

Baseline coverage

Reference methods help show what a model adds beyond simple controls. We track a null control and a conventional method for each protocol.

0 of 2 active baseline roles have published Rewire measurements in this release. Measurements on a selected protocol do not establish coverage of an entire suite.

No execution recipe linked to this protocol. Recipe availability does not establish a completed evaluation.

Author-reported evaluations: 9
External evaluations: 2
paper compilation: 7

Literature evidence is not a Rewire measurement. Executed but unpublished runs and private review status are not included.

Null control

Proposed control: requires review

Select a task-valid null control after reviewing inputs and metric

Protocol-specific applicability, permitted inputs, access, split, evaluator and execution requirements need review before implementation or execution.

This is a suggested selection rule, not a validated method or a measured score.

Conventional reference

Proposed control: requires review

Select an upstream conventional reference after reviewing the full protocol

Protocol-specific applicability, permitted inputs, access, split, evaluator and execution requirements need review before implementation or execution.

This is a suggested selection rule, not a validated method or a measured score.

Protocol coverage CSV (gzip) · Model evaluation matrix (gzip) · Source table (gzip) · Release and checksums (gzip)

Coverage is derived from release 2026-10-23-5347695500ed. Source citations describe the original records; they do not validate an unreviewed baseline proposal. No results have been generated by this audit.

Run instructions

No runnable recipe has been reviewed for this protocol. Dataset access, model requirements, licences and compute requirements must be checked against its sources before execution.

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

MassSpecGym in the Wild: Uncovering and Correcting Evaluation Pitfalls in AI-Driven Molecule Discovery · Original source · arXiv:2606.19624v1 (submitted 2026-06-17), arXiv HTML rendering

Download this release (gzip)

Technical metadata and extraction receipts

Stable ID: lit-otheromics-20261011-msgwild-protocol-formula-retrieval-hit-rate-1

areas: metabolomics
contexts: research
protocol: MassSpecGym formula-known retrieval (bonus); Hit rate@1
version: MassSpecGym as audited in arXiv:2606.19624v1
metric: top-1-accuracy
metric direction: higher
unit: percent
metric qualifier: retrieval
source locator: Table 3 column 'Hit rate@1'; Appendix 'Confidence Intervals'
limitations: Rows mix values copied from the literature with the authors' own runs; see each evaluation's origin.; Grey rows are results the authors consider leaking or inflated.

Related records