Queue
628 cases were harvested and are backed by a real paper, but they fail the test the index publishes on. Each one is listed here with the reason, so the exclusion can be checked rather than trusted.
| Paper | Year | Cited | Why it is held |
|---|---|---|---|
| Modeling Xanthan Gum Foam’s Material Properties Using Machine Learning Metho... | 2024 | 21 | assisted The primary result is the human-produced and human-tested xanthan gum foam material; the ML models only fit and predicted properties of the experimentally measured formulations, serving as an efficiency/analysis component in a human-run experimental workflow. “Five machine-learning-based methods ... were utilized to analyze the effects of the components used in the foam formulation. These models yielded accu...” |
| End-to-end material thermal conductivity prediction through machine learning | 2023 | 21 | unsupported no abstract available for this DOI |
| Forecasting Demand in Supply Chain Using Machine Learning Algorithms | 2015 | 21 | unsupported no abstract available for this DOI |
| Virtual sample generation in machine learning assisted materials design and... | 2023 | 20 | not a result This is a review paper summarizing existing virtual sample generation algorithms and methods, with no new primary result produced. “This review aims to discuss the applications of VSG techniques in machine learning-assisted materials design and discovery based on the research progr...” |
| Prediction of Three-Directional Ground Reaction Forces during Walking Using... | 2023 | 20 | assisted The paper's primary contribution is a human-engineered shoe sole sensor system, with ML regression serving as a calibration/prediction step inside that human-run instrumentation workflow, validated against force plates, rather than an AI capability that itself constitutes the result. “A regression model to predict three-directional ground reaction forces (GRFs) from force sensor outputs was created using multiple linear regression a...” |
| A Survey of Incremental Deep Learning for Defect Detection in Manufacturing | 2024 | 20 | not a result The paper is a survey/review of incremental learning literature for defect detection, presenting no primary result of its own. “This paper reviews how new process, training or validation information is rigorously incorporated in real time when detection exceptions arise during...” |
| Deep learning classification method for boar sperm morphology analysis | 2024 | 20 | not a result The abstract contains only background motivation about the lack of morphology analysis methods, reporting no model, results, or findings from the deep learning approach. “few exist for analyzing morphology outside of subjective manual counting” |
| Automatic improvement of deep learning-based cell segmentation in time-lapse... | 2021 | 20 | assisted The paper's contribution is a human-designed NAS methodology (search space and search strategy) whose resulting networks are benchmarked for accuracy on an existing challenge task, making the AI the method under evaluation in a human-run pipeline rather than the agent whose output constitutes the contribution. “Comprehensive evaluations on the 2D datasets from the cell tracking challenge demonstrate the competitiveness of the proposed method compared to the s...” |
| Evaluation of Machine Learning Algorithms for Air Quality Index (AQI) Predic... | 2023 | 19 | assisted The paper is a human-run comparative evaluation of six standard off-the-shelf ML algorithms reporting accuracy on the existing task of AQI prediction, so the AI is a component whose performance is benchmarked rather than a novel capability the paper contributes. “These machine-learning algorithms have been evaluated based on precision, recall, accuracy, etc. The result shows that Random Forest and Decision Tree...” |
| Machine learning techniques for default prediction: an application to small... | 2023 | 19 | assisted The ML models are evaluated as prediction components within a human-run comparative study of credit risk methods, with the paper's contribution being the performance comparison rather than the AI end-to-end solving default prediction. “This paper uses a large sample of small Italian companies to compare the performance of various machine learning classifiers and a more traditional lo...” |
| Machine learning for combinatorial optimization of brace placement of steel... | 2018 | 19 | assisted The ML component only pre-screens candidate solutions to skip costly structural analyses within a human-designed simulated annealing optimization pipeline, rather than itself producing the optimal brace placement. “the computational cost is successfully reduced by avoiding costly structural analysis for a solution judged by machine learning as nonoptimal” |
| Symbolic regression as a feature engineering method for machine and deep lea... | 2024 | 19 | assisted Symbolic regression is used as a feature-engineering component inside a human-run ML pipeline, with the paper's contribution being a method demonstration of performance gains rather than an AI system itself solving a problem end to end. “we propose to integrate symbolic regression (SR) as an FE process before a ML model to improve its performance” |
| Accelerating process control and optimization via machine learning: a review | 2025 | 18 | unsupported no abstract available for this DOI |
| Quantifying the effects of environment and population diversity in multi-age... | 2022 | 18 | assisted The paper's contribution is a human-designed empirical study quantifying how training factors (procedural generation, population size, intrinsic motivation) affect RL agents' generalization, with the agents serving as subjects of analysis rather than the AI itself producing the solved result. “we investigate and quantify the relationship between generalization and diversity in the multi-agent domain” |
| Using machine learning for transient classification in searches for gravitat... | 2020 | 18 | assisted The ML classifier is one component in a human-run follow-up workflow, producing prioritization information for human telescope-time decisions rather than itself discovering counterparts, and the paper is a method demonstration on real and simulated data. “Towards this end, we use astrorapid, a multiband photometric light-curve classifier, to differentiate between kilonovae, supernovae, and other possibl...” |
| Uncertainty Quantification for Machine Learning‐Based Ionosphere and Space W... | 2023 | 17 | assisted The human authors implement and comparatively analyze uncertainty quantification methods around an ML forecasting model, making the AI a studied component in a human-run methodological evaluation rather than the agent producing the paper's result. “In this paper, we implement and analyze several uncertainty quantification approaches for an ML‐based model to forecast Vertical Total Electron Conten...” |
| Forest Disturbance Monitoring Using Cloud-Based Sentinel-2 Satellite Imagery... | 2024 | 17 | assisted The random forest classifier is one component in a human-run remote-sensing workflow (field inventory data, NDVI index mapping, accuracy validation) whose overall contribution is the monitoring method, not an AI-produced end-to-end result. “a random forest machine learning classifier was applied to show the spatial distribution of dominant species” |
| EssayGAN: Essay Data Augmentation Based on Generative Adversarial Networks f... | 2022 | 17 | assisted EssayGAN is a data-augmentation component whose generated essays feed into a human-built and human-run AES training pipeline, rather than an AI that itself performs the end task of essay scoring. “the generated essays can be used in the supervised learning process for AES systems. Experimental results show that data augmentation using augmented...” |
| Rain Discrimination with Machine Learning Classifiers for Opportunistic Rain... | 2023 | 16 | assisted The paper is a human-designed comparative evaluation of four off-the-shelf ML classifiers as one classification component inside a human-run rain-detection pipeline (data collection, preprocessing, feature extraction all done by the researchers), with the contribution being the classifier comparison rather than an AI capability that itself solved the problem end to end. “It reports a comparison among the results obtained by applying four different machine learning (ML) classifiers, namely the support vector machine (SV...” |
| Enriching automatic test case generation by extracting relevant test inputs... | 2025 | 16 | assisted The LLM is one component inside BRMiner, a human-designed pipeline whose extracted inputs merely feed traditional test generation tools (EvoSuite, Randoop) that actually produce the tests and detect the bugs. “leverages Large Language Models (LLMs) in combination with traditional techniques to extract relevant inputs from bug reports, thereby enhancing autom...” |
| <scp>GeoAnalystBench</scp> : A <scp>... | 2025 | 16 | not a result This is a benchmark/evaluation-framework paper whose contribution is the assessment instrument and findings about model limitations, not an AI system that solved geospatial analysis tasks; the abstract itself calls for evaluation 'before making claims about full GIS automation' and positions the work as a reproducible framework enabling future research. “we present GeoAnalystBench, a benchmark of 50 Python‐based tasks derived from real‐world geospatial problems and validated by GIS experts” |
| Stress-constrained optimization of multiscale structures with parameterized... | 2024 | 15 | assisted The neural networks serve as surrogate models for property and stress prediction inside a human-designed topology optimization framework, so AI is a component in a pipeline humans ran rather than the agent producing the designs. “their material properties are estimated using a neural network surrogate model for homogenized constitutive relations” |
| Usage of Large Language Model for Code Generation Tasks: A Review | 2025 | 15 | not a result This is a literature review surveying 66 papers on LLM-based code generation, not a primary result produced by an AI system. “Using the PRISMA methodology, we reviewed 66 papers published between 2021 and 2023.” |
| Integrating machine learning and symbolic regression for predicting damage i... | 2025 | 15 | assisted The ML and symbolic-regression components were guided tools inside a human-orchestrated framework of specimen fabrication, experimental testing, FEM validation, and design-of-experiments, with the resulting equation serving as a design aid rather than the AI itself delivering the end result. “Feature selection and ML models analyzed the dataset, providing insights that guided PySR in deriving interpretable equations.” |
| Machine learning-driven discovery of novel therapeutic targets in diabetic f... | 2024 | 15 | assisted Machine learning feature selection was one step in a human-run multi-stage bioinformatics pipeline (differential expression, enrichment, ML, independent validation, single-cell analysis) that produced computational gene associations proposed as theoretical targets, not a demonstrated therapeutic result. “Through this integrated approach, SCUBE1 and RNF103-CHMP3 were identified as key genes significantly associated with DFU.” |
| Machine Learning-Based Cost-Effective Smart Home Data Analysis and Forecasti... | 2023 | 14 | assisted The AI model is one component in a human-run workflow (sensor correlation analysis, sensor removal, cost-reduction design); the paper reports a model's RMSE on a prediction task inside a larger design methodology, and its stated contribution is a design guide rather than the AI capability itself. “This process can reduce costs by 13%, thereby providing a design guide for a cost-effective IoT system.” |
| COFFE: A Code Efficiency Benchmark for Code Generation | 2025 | 14 | not a result This is a benchmark/dataset paper whose contribution is an evaluation framework and findings from testing existing LLMs, not an AI system solving a primary problem. “we propose COFFE, a code generation benchmark for evaluating the time efficiency of LLM-generated code solutions” |
| Machine Learning-Assisted Polymer and Polymer Composite Design for Additive... | 2026 | 13 | not a result The abstract explicitly describes a review/survey of ML applications in polymer design for additive manufacturing, with no primary result produced by an AI system. “This review explores how machine learning (ML) is being used to address these challenges.” |
| Non-Perturbative Identification and Subtyping of Amyloidosis in Human Kidney... | 2023 | 13 | assisted The ML (t-SNE/DBSCAN) is a standard analysis component inside a human-run Raman spectroscopy feasibility study on already-diagnosed biopsy samples, and the abstract only claims 'potential' for identification rather than the AI itself delivering a novel solved result. “Our approach, using Raman spectroscopy with machine learning algorithms, shows the potential for the identification of amyloid in pathologic lesions.” |
| LTPNet Integration of Deep Learning and Environmental Decision Support Syste... | 2025 | 12 | unsupported no abstract available for this DOI |
| Controllable Music Playlist Generation Based on Knowledge Graph and Reinforc... | 2022 | 12 | assisted The paper proposes a playlist-generation method and validates it by reporting prediction performance on existing listening-history data, i.e., a model evaluated on an existing recommendation task rather than an AI-produced result that constitutes the contribution. “We confirm the effectiveness of the proposed method by verifying the prediction performance based on listening history and the guidance performance to...” |
| Prediction of Temperature of Liquid Steel in Ladle Using Machine Learning Te... | 2023 | 11 | assisted Machine learning is utilized as one component within a human-developed hybrid modeling workflow (FEM solution of the Fourier equation, inverse parameter identification, production-line verification), so the AI did not itself produce the primary result. “The models were based on the FEM solution of the Fourier equation, and machine learning approaches such as decision trees, linear regression, and arti...” |
| Machine Learning-Based Prediction of EDM Material Removal Rate and Surface R... | 2025 | 11 | unsupported no abstract available for this DOI |
| RetroRanker: leveraging reaction changes to improve retrosynthesis predictio... | 2023 | 10 | assisted RetroRanker is a re-ranking module layered on top of existing retrosynthesis models rather than a system that performs retrosynthesis end to end, and the reported contribution is improved benchmark accuracy for other models' candidate lists. “designed to mitigate the frequency bias in predictions of existing retrosynthesis models through re-ranking” |
| Machine learning-based damage sensing and self-healing of carbon fiber/nylon... | 2023 | 10 | assisted The machine learning predicted damage locations and recommended healing parameters, but the actual self-healing was a physical process (applying electrical current for local heating) carried out by the experimental setup, making the AI an informing component in a human-run materials workflow rather than the agent that produced the healing result. “The self-healing conditions such as the current input pairs of ACN and amount of the electrical current were determined through the artificial neural...” |
| Ontology Completion with Graph-Based Machine Learning: A Comprehensive Evalu... | 2022 | 10 | not a result The paper's contribution is a systematic benchmark evaluation of thirteen existing link-analysis methods on eight existing ontologies, not an AI system solving ontology completion as its primary result. “To the best of our knowledge, this is currently the most extensive systematic study of the applicability of different types of link analysis methods a...” |
| Financial Statement Fraud Detection Through an Integrated Machine Learning a... | 2025 | 10 | assisted The ML models produce fraud-risk predictions and interpretability outputs that inform human audit decisions (framed via decision curve analysis and audit cost simulation), making this a decision-support component evaluated retrospectively rather than an AI end-to-end solving fraud detection. “Decision curve analysis showed that calibrated thresholds improved decision efficiency by 7–9% and reduced over-audit costs by 3–4%, while an audit co...” |
| Interpretable Machine Learning for High-Accuracy Reservoir Temperature Predi... | 2025 | 9 | assisted The paper's contribution is a human-run comparative benchmark of ML models reporting accuracy metrics on an existing prediction task, not an AI-produced result that itself constitutes the solved problem. “This study conducts a comprehensive comparative analysis of advanced machine learning models... to evaluate their predictive performance for reservoir...” |
| Reinforcement learning for mutation operator selection in automated program... | 2025 | 9 | assisted The reinforcement learning agent is only a component (mutation operator selection) inside a heuristic program repair pipeline, and it did not improve the end task of patching bugs compared to random selection. “does not exhibit a noticeable improvement in the number of bugs patched in comparison with the baseline, uniform random selection” |
| Automated Essay Scoring Using Transformer Models | 2021 | 9 | assisted The paper's contribution is a methodological comparison of transformer vs. BOW models on a manually labeled classification task, with the AI positioned as a tool to help human raters rather than an agent that itself solved a problem end to end. “Further, we show how such models can help increase the accuracy of human raters, and we provide a detailed instruction on how to implement transformer...” |