CORE-bench v1.1
Collection
Benchmark for AI agents on scientific reproducibility — mainline (39) and OOD (19) splits derived from Code Ocean capsules. • 2 items • Updated
field stringclasses 3
values | language stringclasses 2
values | capsule_title stringlengths 17 127 | capsule_id stringlengths 15 15 | task_prompt stringlengths 14 410 | gpu bool 1
class | results listlengths 3 3 | capsule_doi stringlengths 38 64 |
|---|---|---|---|---|---|---|---|
Computer Science | Python | Detection and State Analysis of Drowsiness using Multitask Learning with Neural Networks | capsule-5507257 | Run multiclass_state_analysis_testing.py | true | [
{
"Report the test accuracy of 'mouth talking'.": 96.1249913532
},
{
"Report the test accuracy of 'mouth talking'.": 96.1249913532
},
{
"Report the test accuracy of 'mouth talking'.": 96.1249913532
}
] | https://doi.org/10.24433/CO.0217715.v1 |
Computer Science | Python | Short-Term Temperature Forecasts Using a Convolutional Neural Network | capsule-3449234 | Run the jupyter notebook visualize_results.ipynb using a python3 kernel and convert it to html. For all the runs, disable the cell execution timeout and allow errors. | true | [
{
"fig From the figure containing the standard deviation for the Essen data, report the name of the model with the highest standard deviation between time 0 and 10.": "SN"
},
{
"fig From the figure containing the standard deviation for the Essen data, report the name of the model with the highest standa... | https://doi.org/10.24433/CO.8788349.v1 |
Computer Science | Python | Network Diffusion Examples | capsule-8807709 | Run 'epidemic.py'. | null | [
{
"fig For the third subplot in the visualization of the experiments, report the color of the line with the greatest number of nodes at epoch 15.": "blue",
"fig Report the name of the first subplot in the visualization of the experiments.": "ill"
},
{
"fig For the third subplot in the visualization ... | https://doi.org/10.24433/CO.1013089.v4 |
Computer Science | Python | Multi-Instance Ensemble Learning with Discriminative Bags (ELDB) | capsule-6049678 | Run 'Main.py'. | null | [
{
"Report the f1 score for the Musk1+ dataset with the svm classifier.": 87.516,
"Report the f1 score for the Musk1+ dataset with the j48 classifier.": 79.813
},
{
"Report the f1 score for the Musk1+ dataset with the svm classifier.": 86.578,
"Report the f1 score for the Musk1+ dataset with the ... | https://doi.org/10.24433/CO.1490343.v1 |
Computer Science | R | mFLICA Reproducible Capsule | capsule-2804717 | Export the following R default packages: datasets,utils,grDevices,graphics,stats,methods. Run 'ResultReproducibilityNotebook.Rmd' and render it as an html file. Store the output in ../results and set clean to 'TRUE'. | null | [
{
"fig From the Network Density plot (Figure 4), report the label for the red line (ignore spaces).": "TS#1"
},
{
"fig From the Network Density plot (Figure 4), report the label for the red line (ignore spaces).": "TS#1"
},
{
"fig From the Network Density plot (Figure 4), report the label for th... | https://doi.org/10.24433/CO.4248204.v1 |
Computer Science | Python | Explainable Machine Learning Pipeline for Twitter Bot Detection | capsule-3418007 | Run 'main.py'. | null | [
{
"Report the F1 score for statistical general only.": 0.7536528882,
"fig Report the proposed model's AUC from the ROC curves figure. Ignore the confidence interval.": 0.98
},
{
"Report the F1 score for statistical general only.": 0.75695913,
"fig Report the proposed model's AUC from the ROC cur... | https://doi.org/10.24433/CO.9672787.v1 |
Computer Science | Python | Feature Selection | capsule-1624349 | Execute 'FS-Filters.ipynb'. Save the results in html format in ../results. For all the runs, disable the cell execution timeout and allow errors. | null | [
{
"Report the best accuracy of the hybrid filter wrapper strategy.": 0.9506166188,
"fig Report the name of the feature with the highest I-Gain.": "RAWRED-MEAN"
},
{
"Report the best accuracy of the hybrid filter wrapper strategy.": 0.9506166188,
"fig Report the name of the feature with the highe... | https://doi.org/10.24433/CO.1509005.v1 |
Social Sciences | R | Reproducible research practices and transparency across linguistics | capsule-9832712 | Create the following three directories in the results folder: 01_scopus-selection, 02_coding, 03_analyses. Run 'master_script.R' using Rscript. | null | [
{
"fig From Figure 2 in the cleaned results, report the percentage of 'Not Available' analysis scripts for Pre-RC (2008-09)": 100,
"fig From Table 1 in the cleaned results, report the number of Included Articles After.OS": 255
},
{
"fig From Figure 2 in the cleaned results, report the percentage of ... | https://doi.org/10.24433/CO.2289033.v2 |
Medical Sciences | R | ESR1 mutant breast cancers show elevated basal cytokeratins and immune activation | capsule-2816027 | Export the following R default packages: datasets,utils,grDevices,graphics,stats,methods. Then, run 'main.R' using Rscript. | null | [
{
"fig For CTCF Signature Enrichment, report the name of the group with the highest median GSVA score.": "MCF7_D538G"
},
{
"fig For CTCF Signature Enrichment, report the name of the group with the highest median GSVA score.": "MCF7_D538G"
},
{
"fig For CTCF Signature Enrichment, report the name ... | https://doi.org/10.24433/CO.0627595.v1 |
Social Sciences | R | Adaptations to sea level change and transitions to agriculture at Khao Toh Chong rockshelter, Peninsular Thailand | capsule-3821950 | Create a 'figures' directory in the results folder. Run 'ktc_11_paper.Rmd' using Rscript and render it as html. Save the output to the ../results output directory. Set clean to 'TRUE'. | null | [
{
"fig Report the name of the material with the highest Depth Below Surface at 10,000 calibrated years BP.": "charcoal",
"fig Report the name of the material with the highest mass (g) and 5000 years cal. BP.": "ceramics"
},
{
"fig Report the name of the material with the highest Depth Below Surface ... | https://doi.org/10.24433/CO.0c427240-6e28-417d-ba1a-777c8c4e485a |
Medical Sciences | R | Framework to study harms and benefits of multi-cancer early detection tests | capsule-9054015 | Run 'pancancer_calculation.R' using Rscript. | null | [
{
"fig Report the percentage sensitivity for cancers A and B that has the highest number of cancers detected per 1,000 women for a 1.0% prevalance of cancer B.": 90
},
{
"fig Report the percentage sensitivity for cancers A and B that has the highest number of cancers detected per 1,000 women for a 1.0% ... | https://doi.org/10.24433/CO.4448554.v3 |
Computer Science | Python | Deep Learning for Cellular Traffic Prediction | capsule-3301293 | Run 'run_prediction.py'. | true | [
{
"Report the test RMSE of the model.": 26.212040000000002
},
{
"Report the test RMSE of the model.": 26.212040000000002
},
{
"Report the test RMSE of the model.": 26.212040000000002
}
] | https://doi.org/10.24433/CO.761a91d4-6f2f-4912-8129-2bb8abeaa044 |
Social Sciences | R | Adapting the coordination of eyes and head to differences in task and environment during fully-mobile visual exploration | capsule-1724988 | Run 'calibration_error.R', 'lss1_summary_analyses.R', 'lss2_summary_analyses.R', and 'lss2_peak_analyses.R' all using Rscript. | null | [
{
"fig Report the task name with the higher median walking speed (m/s).": "Walk",
"fig Report the task name with the higher median straightness ratio.": "Search"
},
{
"fig Report the task name with the higher median walking speed (m/s).": "Walk",
"fig Report the task name with the higher median ... | https://doi.org/10.24433/CO.8767371.v2 |
Social Sciences | R | Preaching to the Choir: A Problem of Participatory Interventions | capsule-4299879 | Run '01_motivation.R', '02_design.R', '03_survey.R', '04_metaketa_comp.R', '05_lapop.R', '06_misc.R' using Rscript. | null | [
{
"fig From the figure measuring homicide rate per 100k in the last 12 months, report the name of the sample with the lower homicide rate per 100k in 2000.": "Colombia"
},
{
"fig From the figure measuring homicide rate per 100k in the last 12 months, report the name of the sample with the lower homicide... | https://doi.org/10.24433/CO.0479587.v1 |
Medical Sciences | Python | MLP-based classification of COVID-19 and skin diseases | capsule-0851068 | Run the bash script 'demo.sh'. | true | [
{
"Report the final AUC after training.": 0.9157952669
},
{
"Report the final AUC after training.": 0.9157952669
},
{
"Report the final AUC after training.": 0.9157952669
}
] | https://doi.org/10.24433/CO.9705378.v1 |
Medical Sciences | R | Compute capsule for stochastic process based COVID-19 simulation environment | capsule-1394704 | Run 'modular.Rmd' using Rscript and render it as a html. Store the output in ../results. Set clean to 'TRUE'. | null | [
{
"fig Report the name of the method with the higher R0.": "ML",
"Report the R0 of EG.": 1.035213
},
{
"fig Report the name of the method with the higher R0.": "ML",
"Report the R0 of EG.": 1.035213
},
{
"fig Report the name of the method with the higher R0.": "ML",
"Report the R0 of... | https://doi.org/10.24433/CO.7958703.v1 |
Social Sciences | R | No Evaluative Conditioning Effects with Briefly Presented Stimuli | capsule-0504157 | Run 'manuscript.Rmd' using Rscript and render it as a pdf. Store the output in the ../results directory. Set clean to 'TRUE'. | null | [
{
"fig From figure 1, report the CS presentation time with the greatest mean EC effect (ignore units).": 1000
},
{
"fig From figure 1, report the CS presentation time with the greatest mean EC effect (ignore units).": 1000
},
{
"fig From figure 1, report the CS presentation time with the greates... | https://doi.org/10.24433/CO.26389ff0-ea56-467d-a550-b96cb9a31e04 |
Computer Science | Python | On the Energy Footprint of Mobile Testing Frameworks | capsule-3593259 | Run 'physalia_automators.reports' as a python module with /results as the output directory. | null | [
{
"fig From the violin plot of the energy comsumption of tap, report the name of the framework that consumes the most energy.": "Appium"
},
{
"fig From the violin plot of the energy comsumption of tap, report the name of the framework that consumes the most energy.": "Appium"
},
{
"fig From the ... | https://doi.org/10.24433/CO.4277516.v1 |
Social Sciences | Python | When to retrieve and encode episodic memories: a neural network model of hippocampal-cortical interaction | capsule-3639589 | Run demo.py in the code/src folder. | null | [
{
"fig Report the color of the line with the highest maximum activation for target memory activation, DM.": "blue"
},
{
"fig Report the color of the line with the highest maximum activation for target memory activation, DM.": "blue"
},
{
"fig Report the color of the line with the highest maximum... | https://doi.org/10.24433/CO.5779179.v3 |
Social Sciences | R | Attentional fluctuations and the temporal organization of memory - Jayakumar, Balusu, & Aly (2023). Cognition. | capsule-2345790 | Set up the following subfolders in the ../results directory: intermediates, figures, stats_figures_markdowns. Run all the .Rmd files using Rscript and render them as html. Store the output files in ../results/stats_figures_markdowns. | null | [
{
"From Study 1, report the mean of the response rate across all participants.": 97.82,
"From Study 2, report the mean of the response rate across all participants.": 99.41
},
{
"From Study 1, report the mean of the response rate across all participants.": 97.82,
"From Study 2, report the mean o... | https://doi.org/10.24433/CO.3162457.v1 |
Medical Sciences | R | Excess significance and power miscalculations in neurofeedback research | capsule-7716865 | Run 'manuscript.Rmd' using Rscript and render it as a pdf. Store the output in ../results. Set clean to 'TRUE'. | null | [
{
"fig From Table 1, report the sensitivity for 80% power and recalculated mean (regulation).": 1.31
},
{
"fig From Table 1, report the sensitivity for 80% power and recalculated mean (regulation).": 1.31
},
{
"fig From Table 1, report the sensitivity for 80% power and recalculated mean (regulat... | https://doi.org/10.24433/CO.7282505.v1 |
Medical Sciences | R | A global and integrated analysis of CINSARC-associated genetic defects | capsule-4933686 | Run "Main.R" using Rscript and xvfb-run. | null | [
{
"fig From Figure 2 plot A, report Fisher's P. If the value is in scientific notion, convert it to a floating point number.": 0.0182,
"fig From Figure 1 plot A, measuring time vs. metastasis-free survival, report the numerical value of HR (ignore the confidence interval).": 2.15
},
{
"fig From Figu... | https://doi.org/10.24433/CO.9777456.v4 |
Social Sciences | R | Automated Text Classification of News Articles: A Practical Guide | capsule-9240688 | Run the bash script 'run.sh'. | null | [
{
"From table 1, report the portion relevant in both corpora.": 0.44
},
{
"From table 1, report the portion relevant in both corpora.": 0.44
},
{
"From table 1, report the portion relevant in both corpora.": 0.44
}
] | https://doi.org/10.24433/CO.4630956.v1 |
Medical Sciences | R | Cardiac structure and function in schizophrenia | capsule-1175539 | Run "/code/CardioSCZ.R" using Rscript. | null | [
{
"fig Report the name of the patient group with the greater median concentricity.": "SCZ"
},
{
"fig Report the name of the patient group with the greater median concentricity.": "SCZ"
},
{
"fig Report the name of the patient group with the greater median concentricity.": "SCZ"
}
] | https://doi.org/10.24433/CO.9265392.v1 |
Medical Sciences | R | Estimating the prevalence of discrepancies between study registrations and publications: A systematic review and meta-analyses | capsule-2708693 | Run 'preregSR_manuscript.Rmd' and render it as a pdf. Store the output in ../results. Set clean as 'TRUE'. | null | [
{
"fig From table 1, report the k value for the medicine discipline.": 81,
"fig From table 3, report n for 'percentage of studies with at least one outcome discrepancy that disclose an outcome discrepancy'.": 620
},
{
"fig From table 1, report the k value for the medicine discipline.": 81,
"fig ... | https://doi.org/10.24433/CO.4753181.v1 |
Medical Sciences | R | Integrative Cancer Pharmacogenomics to Infer Large-Scale Drug Taxonomy [PMID: 28314784] | capsule-4252248 | Create the symbolic links for ../results output. Create the symbolic links for ../data Data. Run 'main-ctrpv.R', 'main-nci.R', and 'main-network-generation.R' using Rscript. | null | [
{
"fig Report the overall AUC from the PR curve generated with the CTRPv2 sensitivity dataset, tested against ATC annotations and drug-target information from CHEMBL.": 0.4929241
},
{
"fig Report the overall AUC from the PR curve generated with the CTRPv2 sensitivity dataset, tested against ATC annotati... | https://doi.org/10.24433/CO.0012b3fb-3cf2-41fa-9b8d-6cc055b53ca2 |
Computer Science | Python | SPT: Security Policy Translator for Network Security Functions in Cloud-Based Security Services | capsule-4671827 | Execute 'PerformanveEval.ipynb'. Save the results in html format in ../results. For all the runs, disable the cell execution timeout and allow errors. | null | [
{
"fig Report the name of the mapping with the higher execution time at 44 elements.": "Semantic-based"
},
{
"fig Report the name of the mapping with the higher execution time at 44 elements.": "Semantic-based"
},
{
"fig Report the name of the mapping with the higher execution time at 44 element... | https://doi.org/10.24433/CO.7409799.v1 |
Medical Sciences | R | lab: An R package for generating analysis-ready data from laboratory records | capsule-7186268 | Run 'SampleCode.Rmd' using Rscript and render it as html. Store the output in ../results. Set clean as 'TRUE'. | null | [
{
"fig From laboratory test 18262-6, report the name of the method with the higher missing rate at gap 30.": "By Window",
"fig From laboratory test 2160-0 Creatinine, report the ID number with the highest laboratory result at window 2.": 109
},
{
"fig From laboratory test 18262-6, report the name of... | https://doi.org/10.24433/CO.9692445.v1 |
Social Sciences | R | Partisan Enclaves and Information Bazaars: Mapping Selective Exposure to Online News | capsule-5136217 | Make the following subfolders in the ../results directory: tables, figures, for_publication/tables, for_publication/figures. Run all the .R scripts in the ../code folder using Rscript with 'source' and set echo to 'TRUE'. | null | [
{
"fig From figure 3 from the figures for publication, report the name of the party ID with the lowest share of political news from portals.": "Lean DEM"
},
{
"fig From figure 3 from the figures for publication, report the name of the party ID with the lowest share of political news from portals.": "Lea... | https://doi.org/10.24433/CO.1889895.v1 |
Medical Sciences | Python | A Predictive Analytics Framework for Early-Stage Thyroid Cancer Using ML | capsule-7800694 | Execute 'Thyroid Cancer Final Code.ipynb'. Save the results in html format in ../results. For all the runs, disable the cell execution timeout and allow errors. | null | [
{
"Which model has the highest Macro F1?": "CatBoost",
"What is the ROC AUC (macro OvR) of the best performing model?": 0.9570000000000001
},
{
"Which model has the highest Macro F1?": "CatBoost",
"What is the ROC AUC (macro OvR) of the best performing model?": 0.9570000000000001
},
{
"W... | https://doi.org/10.24433/CO.2418823.v1 |
Social Sciences | Python | The Sovereign Constitution Fabric (SNF) v2.0 — Reproducible Synthetic-Floor Capsule | capsule-8412128 | Run 'run_all.sh'. | null | [
{
"What is the mean of the snf_effectiveness_index?": 0.5791337825,
"What is the median of the dignity_score?": 0.5888590853
},
{
"What is the mean of the snf_effectiveness_index?": 0.5791337825,
"What is the median of the dignity_score?": 0.5888590853
},
{
"What is the mean of the snf_e... | https://doi.org/10.24433/CO.9005626.v1 |
Computer Science | Python | DignityProof — Global Verifiability Capsule (v1.0) | capsule-7655932 | Run 'run_all.sh'. | null | [
{
"What is the point estimate (deaths averted synthetic) from the child wasting impact model? Give results as integer.": 280000,
"What is the child wasting impact model reported 95% interval (synthetic)? Give a list of 2 integer value for the confidence interval.": [
196000,
364000
]
},
... | https://doi.org/10.24433/CO.0436753.v1 |
Medical Sciences | R | Unravelling the transcriptome of the human tuberculosis lesion and its clinical implications | capsule-9294029 | Run 'script1.R' using Rscript. | null | [
{
"fig what location has the lowest median enrichment score?": "H"
},
{
"fig what location has the lowest median enrichment score?": "H"
},
{
"fig what location has the lowest median enrichment score?": "H"
}
] | https://doi.org/10.24433/CO.4427270.v1 |
Medical Sciences | R | custom code for: Predicting IVF live birth probabilities using live machine learning, center-specific models | capsule-6295990 | Run 'R_demo.R' using Rscript. | null | [
{
"What is the LBP model ROC-AUC (weighted)": 0.7142856999999999,
"What is the percent AUC-Improvement? Give an integer": 25
},
{
"What is the LBP model ROC-AUC (weighted)": 0.7142856999999999,
"What is the percent AUC-Improvement? Give an integer": 25
},
{
"What is the LBP model ROC-AUC... | https://doi.org/10.24433/CO.8413662.v1 |
Medical Sciences | R | VascularBC-ST: the code to analyze the Vascular microenvironment of Breast Cancer using Spatial Transcriptome | capsule-9477017 | Export the following R default packages: datasets,utils,grDevices,graphics,stats,methods. Run 'main.R' using Rscript. | null | [
{
"fig Pearson correlation coefficients between the estimated proportions of different cell types were calculated, what is the highest pearson correlation related to? Give response in a list of strings": [
"Smooth.1",
"Endo.1"
]
},
{
"fig Pearson correlation coefficients between the esti... | https://doi.org/10.24433/CO.9243235.v1 |
Computer Science | Python | Codex Wall v4.0 — Encrypted Holo-Causality & Predictive Governance (Reproducible Capsule) | capsule-0201673 | Run 'run_all.sh'. | null | [
{
"What is the mean_gdp_gain_pct?": 0.9999005881,
"What is the p95?": 1.1643528158,
"What is the p50?": 0.9999966484
},
{
"What is the mean_gdp_gain_pct?": 0.9999005881,
"What is the p95?": 1.1643528158,
"What is the p50?": 0.9999966484
},
{
"What is the mean_gdp_gain_pct?": 0.99... | https://doi.org/10.24433/CO.4072743.v1 |
Medical Sciences | Python | Deep learning-enabled accurate assessment of gait impairments in Parkinson's disease using smartphone videos | capsule-0152700 | Check GPU status using nvidia-smi, run the training, validation, and testing shell scripts (train.sh, val.sh, test.sh), and execute the Python scripts: directory_features_process.py, grad_cam.py, joint_feature_severity_spearman_calculation.py, sig_cal.py, joint_significance_cal_medication_response.py, summary_of_joint_... | null | [
{
"fig are there more than 2 features with accuracy exceeding 0.4? Respond with yes or no.": "yes",
"fig which joint has the lowest absolute spearman correlation for acceleration?": "ankle",
"Given the Kruskal-Wallis for Group 0-2 (Group 1 vs. Group 3), what is the p-value?": "2.341092434893948e-10"
}... | https://doi.org/10.24433/CO.7700825.v1 |
Medical Sciences | Python | A motif preferred adenine base editor with minimal bystander and off-targets editing | capsule-2242462 | Run 'main.py'. | null | [
{
"What is the Chromosome chr1 with positioning 1013983, IsNGGMatch status?": "True",
"What is the Chromosome chr1 with positioning 1319313, ABE8e_NG status?": "False"
},
{
"What is the Chromosome chr1 with positioning 1013983, IsNGGMatch status?": "True",
"What is the Chromosome chr1 with posit... | https://doi.org/10.24433/CO.8003706.v2 |
Computer Science | Python | PM-Score-Net: A Disjoint-Label Multi-Task Learning Framework for Automated Continuous Severity Assessment of Myopic Maculopathy | capsule-3762736 | Run the evaluation scripts: run_baseline_eval.sh, run_convnext_eval.sh, run_scl_eval.sh, run_task1_eval.sh, run_palm_eval.sh, and run_rigor_eval.sh. | null | [
{
"What is the Task1_Only macro f1?": 0.7012,
"What is the Winner_SCL RMSE?": 0.4708
},
{
"What is the Task1_Only macro f1?": 0.7012,
"What is the Winner_SCL RMSE?": 0.4708
},
{
"What is the Task1_Only macro f1?": 0.7012,
"What is the Winner_SCL RMSE?": 0.4708
}
] | https://doi.org/10.24433/CO.6713286.v2 |
No dataset card yet
- Downloads last month
- 53