Back to results

CompBioBench item

ml-model-track-overlap-q1

Question

Borzoi (Linder et al., doi:10.1038/s41588-024-02053-6) and Sei (Chen et al., doi:10.1038/s41588-022-01102-2) are both DNA sequence models trained to predict genomic assay outputs. Out of the Sei tracks that are linked to a specific Cistrome ID, how many tracks share provenance with any of the tracks in Borzoi? Round your answer to the nearest 100.

Inferred answer
3600Uniquely inferred from submitted scores.
Domain
Epigenomics
Question style
Retrieval

An answer can be uniquely inferred for 54 of 100 questions from submitted scores. Green: correct · Red: incorrect. For questions without a uniquely determined answer: Blue: matches prediction · Orange: differs from prediction. These are inferred answer checks, not official per-question grades.

Galaxy

12/12 terminal item results

UDT = user-defined tool: custom code executed by Galaxy. Route badges summarize analysis tools, excluding uploads and automatic conversions. How usage is counted.

Codex GPT-5.5

3 replicates
Codex GPT-5.5

Replicate 1

IncorrectUDT-only
Answer1300
Workflow
Sei Borzoi Provenance Match
Total tokens
2.1M
Output
9.5K
Codex GPT-5.5

Replicate 2

RerunIncorrectUDT-only
Answer1300
Workflow
Cistrome overlap analysis for Sei and BorzoiCistrome old API overlap validation for Sei and BorzoiCistrome single keyword query
Total tokens
2.5M
Output
21.2K
Codex GPT-5.5

Replicate 3

IncorrectUDT + standard tools
Answer100
Workflow
Inspect Sei and Borzoi source annotationsInspect Sei and Borzoi source annotations v2Inspect Sei and Borzoi source annotations v3Minimal path render testMinimal path render test v25 more tools
Total tokens
7.0M
Output
21.3K

Codex GPT-5.6 Sol

3 replicates
Codex GPT-5.6 Sol

Replicate 1

IncorrectUDT-only
Answer2,400
Workflow
Sei Borzoi provenance diagnosticSei Borzoi exact provenance analysisSei Borzoi exact provenance analysis v2Sei Borzoi provenance by Sei Cistrome records
Total tokens
22.9M
Output
29.8K
Codex GPT-5.6 Sol

Replicate 2

IncorrectUDT + standard tools
Answer2400
Workflow
Sei Borzoi provenance diagnosticSei Borzoi provenance diagnostic v2Sei Borzoi provenance diagnostic v3UDT copy probeSei Borzoi provenance diagnostic v46 more tools
Total tokens
16.2M
Output
31.7K
Codex GPT-5.6 Sol

Replicate 3

IncorrectUDT-only
Answer2400
Workflow
Borzoi Sei provenance diagnosticBorzoi Sei provenance diagnostic revisedBorzoi Sei shared provenance final count
Total tokens
6.5M
Output
16.7K

Codex + DeepSeek-v4-pro-0813

3 replicates
Codex + DeepSeek-v4-pro-0813

Replicate 1

IncorrectUDT-only
Answer1300
Workflow
Sei Borzoi Provenance CompareSei Borzoi Provenance Compare v2
Total tokens
6.4M
Output
36.8K
Codex + DeepSeek-v4-pro-0813

Replicate 2

IncorrectUDT-only
Answer2400
Workflow
Borzoi Sei Cistrome Provenance OverlapBorzoi Sei Cistrome Provenance Overlap v3Borzoi Sei Cistrome Provenance Overlap v4
Total tokens
22.0M
Output
59.1K
Codex + DeepSeek-v4-pro-0813

Replicate 3

CorrectUDT-only
Answer3600
Workflow
Probe Cistrome APIResolve Sei Cistrome IDsResolve Sei Cistrome IDs v2Inspect Sei Cistrome MappingResolve Borzoi Provenance4 more tools
Total tokens
14.4M
Output
73.4K

Codex GPT-5.6 Luna

3 replicates
Codex GPT-5.6 Luna

Replicate 1

IncorrectUDT-only
Answer100
Workflow
Track provenance diagnosticCistrome/Borzoi provenance overlap diagnosticFinal Sei/Borzoi provenance overlap
Total tokens
7.1M
Output
20.4K
Codex GPT-5.6 Luna

Replicate 2

IncorrectUDT + standard tools
Answer100
Workflow
Xlsx2TsvFilter TabularSei Borzoi provenance diagnosticUDT render probeUDT command probe6 more tools
Total tokens
36.1M
Output
92.4K
Codex GPT-5.6 Luna

Replicate 3

IncorrectUDT + standard tools
Answer100
Workflow
Tp Grep ToolUpload1Xlsx2TsvSei Borzoi provenance probeUDT render probe6 more tools
Total tokens
21.5M
Output
61.3K

Anycode

13/13 terminal item results

Codex GPT-5.5

3 replicates
Codex GPT-5.5

Replicate 1

Correct
Answer3600
Workflow
Local codePython
Codex GPT-5.5

Replicate 2

Incorrect
Answer2400
Workflow
Local codePython
Codex GPT-5.5

Replicate 3

Correct
Answer3600
Workflow
Local codePython

Codex GPT-5.6 Sol

3 replicates
Codex GPT-5.6 Sol

Replicate 1

Incorrect
Answer1300
Workflow
Local codePython
Codex GPT-5.6 Sol

Replicate 2

Incorrect
Answer1300
Workflow
Local codePython
Codex GPT-5.6 Sol

Replicate 3

Correct
Answer3600
Workflow
Local codePython

Codex + DeepSeek-v4-pro-0813

3 replicates
Codex + DeepSeek-v4-pro-0813

Replicate 1

Incorrect
Answer3400
Workflow
Local codePython
Total tokens
51.9M
Output
99.0K
Codex + DeepSeek-v4-pro-0813

Replicate 2

Incorrect
Answer1300
Workflow
Local codePython
Total tokens
13.8M
Output
50.8K
Codex + DeepSeek-v4-pro-0813

Replicate 3

Correct
Answer3600
Workflow
Local codePython
Total tokens
38.5M
Output
130.5K

Codex GPT-5.6 Luna

3 replicates
Codex GPT-5.6 Luna

Replicate 1

Incorrect
Answer100
Workflow
Local codePython
Codex GPT-5.6 Luna

Replicate 2

Incorrect
Answer1300
Workflow
Local codePython
Codex GPT-5.6 Luna

Replicate 3

Incorrect
Answer1300
Workflow
Local codePython

Codex GPT-6 Astra

1 replicate
Codex GPT-6 Astra

Replicate 1

Correct
Answer3600
Workflow
Local codePython
Total tokens
1.0M
Output
6.4K