In the rapidly evolving domain of Recommender Systems (RecSys), new algorithms frequently claim state-of-the-art performance based on evaluations over a limited set of arbitrarily selected datasets. However, this approach may fail to holistically reflect their effectiveness due to the significant impact of dataset characteristics on algorithm performance. Addressing this deficiency, this paper introduces a novel benchmarking methodology to facilitate a fair and robust comparison of RecSys algorithms, thereby advancing evaluation practices. By utilizing a diverse set of 30 open datasets, including two introduced in this work, and evaluating 11 collaborative filtering algorithms across 9 metrics, we critically examine the influence of dataset characteristics on algorithm performance. We further investigate the feasibility of aggregating outcomes from multiple datasets into a unified ranking. Through rigorous experimental analysis, we validate the reliability of our methodology under the variability of datasets, offering a benchmarking strategy that balances quality and computational demands. This methodology enables a fair yet effective means of evaluating RecSys algorithms, providing valuable guidance for future research endeavors.

High-dimensional tabular data are common in biomedical and clinical research, yet conventional machine learning methods often struggle in such settings due to data scarcity, feature redundancy, and limited generalization. In this study, we systematically evaluate Synolitic Graph Neural Networks (SGNNs), a framework that transforms high-dimensional samples into sample-specific graphs by training ensembles of low-dimensional pairwise classifiers and analyzing the resulting graph structure with Graph Neural Networks. We benchmark convolution-based (GCN) and attention-based (GATv2) models across 15 UCI datasets under two training regimes: a foundation setting that concatenates all datasets and a dataset-specific setting with macro-averaged evaluation. We further assess cross-dataset transfer, robustness to limited training data, feature redundancy, and computational efficiency, and extend the analysis to a real-world ovarian cancer proteomics dataset. The results show that topology-aware node feature augmentation provides the dominant performance gains across all regimes. In the foundation setting, GATv2 achieves an ROC-AUC of up to 92.22 (GCN: 91.22), substantially outperforming XGBoost (86.05), $\alpha =0.001$. In the dataset-specific regime, GATv2, combined with minimum-connectivity filtering, achieves a macro ROC-AUC of 83.12, compared to 80.28 for XGBoost. Leave-one-dataset-out evaluation confirms cross-domain transfer, with an ROC-AUC of up to 81.99. SGNNs maintain ROC-AUC around 85% with as little as 10% of the training data and consistently outperform XGBoost in more extreme low-data regimes, $\alpha =0.001$. On ovarian cancer proteomics data, foundation training improves both predictive performance and stability. Efficiency analysis shows that graph filtering substantially reduces training time, inference latency, and memory usage without compromising accuracy. Overall, these findings suggest that SGNNs provide a robust and scalable approach for learning from high-dimensional, heterogeneous tabular data, particularly in biomedical settings with limited sample sizes.

In this study, we present a systematic evaluation of Synolitic Graph Neural Networks (SGNNs), a novel framework that transforms high-dimensional tabular data into sample-specific graphs using ensembles of low-dimensional pairwise classifiers. We demonstrate that augmenting these graphs with topology-aware node descriptors (such as degree, strength, closeness, and betweenness centrality) and applying graph sparsification techniques, either via minimum spanning connectivity or fixed-probability edge retention, can significantly improve classification performance. We evaluate both convolution-based (GCN) and attention-based graph neural networks (GATv2) across two training regimes: a foundation model setting where multiple datasets are concatenated, and dataset-specific training. Results show that attention-based models generally exhibit superior performance across classification tasksвАФin the foundation regime, dense (non-sparsified) graphs with node features yield 92.83 ROC-AUC for GATv2 and 92.34 for GCN (vs. 90.80 for XGBoost), and in the dataset-specific regime, GATv2 with minimal connectivity and node features reaches 88.96 ROC-AUC (vs. 86.84 for XGBoost). A leave-one-dataset-out evaluation further indicates out-of-domain transfer to previously unseen datasets (mean ROC-AUC: 0.78 with node features; 0.71 with maximum-threshold sparsification; 0.70 without features). Importantly, we demonstrate that the SGNN framework is capable of overcoming the curse of dimensionality, outperforming traditional machine learning models such as XGBoost in scenarios where the number of features exceeds the number of training samplesвАФmaintaining ROC-AUC above 80\% with only 5\% of the training data, while XGBoost drops to 60\%. Furthermore, SGNNs exhibit robustness to feature redundancy and correlation, with duplicating all features and adding noise producing only minor deviations, reducing the need for manual feature engineering or dimensionality reduction. Across all settings, SGNNs enhanced with node features consistently outperform XGBoost baselines, underscoring the effectiveness of integrating graph-based structural representations, topology-aware augmentation, and controlled sparsification in classification tasks.

Large language model (LLM) agents improve task performance by decomposing problems into role-specialized behaviors. However, their practical deployment is often limited by the computational cost and instability associated with the on-the-fly agent design for each user request. To address this, we present LLM Agents Factory, a retrieval-based framework that constructs domain-specific and Wikipedia-grounded agents on demand using a base of over 20K predetermined agent profiles. Our framework supports two modes: (1) agent profile retrieval via semantic search and (2) distillation into a compact model fine-tuned for direct agent generation. Experiments on MMLU, BIG-bench, and BIG-bench Hard in a single-agent scenario demonstrate that our retrieval-based agent construction surpasses non-agent baselines in accuracy while matching AutoGen generation quality with a 120B backbone at a substantially lower inference cost. Our work reveals that retrieval from a structured agent repository provides a cost-efficient, accurate, and controllable alternative to dynamic agent generation, responding to the strict demands of industrial applications. We provide the implementation code and the agent base in https://huggingface.co/frontier-ai/llm-agent-factory.