Skill UI

Collection of battle-tested PyTorch training recipes covering architecture selection, optimizers, LR scheduling, precision tips, and debugging to train or fine-tune LLMs, vision, diffusion, and biomedical models with high GPU throughput.

A comprehensive expert prompt for evaluating Power BI data models. It guides the assessment of model architecture (Star Schema, relationships), storage optimization (Import, DirectQuery, Dual), data quality, performance scalability, and governance best practices. Use this when reviewing an existing BI model to ensure it is robust, maintainable, and highly performant.

This guide provides a comprehensive reference architecture for building production-ready, enterprise-grade applications on Palantir Foundry. It covers the entire data lifecycle, detailing best practices for the standard data pipeline (Raw > Clean > Model), Ontology design, external API integration patterns, and robust multi-layered security models. Use it when designing, planning, or optimizing complex data infrastructure within Foundry.

This skill provides a structured framework for designing and implementing composable recommendation, ranking, and feed pipelines. It models the industry-standard six-stage pattern: Source, Hydrator, Filter, Scorer, Selector, and SideEffect. Use it for any system that selects "the top K items for a (user, context)", such as social feeds, ad ranking, RAG rerankers, or content recommendation engines, ensuring a robust and scalable architecture.

A comprehensive tool for training and evaluating advanced neural network models (including LSTM, Transformer, and N-BEATS) on time-series market data. It facilitates the prediction of financial trends, generating results with confidence intervals, and comparing performance across multiple architectures. Ideal for quantitative trading research and financial forecasting.

This skill utilizes hyperbolic space embedding (Poincare ball model) to represent complex data structures with inherent hierarchy, such as dependency trees, ontologies, or module taxonomies. Unlike standard Euclidean embeddings, hyperbolic space preserves hierarchical distances efficiently, allowing for accurate distance calculations (geodesic distance) that grow logarithmically with tree depth. It is ideal for advanced knowledge graph analysis and code architecture mapping.

A comprehensive guide for designing conceptual figures for highly theoretical manuscripts (e.g., Psychological Review). This skill teaches how to create two types of non-experimental exhibits: structured model/architecture diagrams and simulation-of-model-behavior plots. It emphasizes rigorous academic standards, including defining constructs vs. processes, distinguishing predictions from fits, and ensuring figures are self-contained and diagnostic.

A comprehensive guide detailing the rigorous, multi-stage academic review process for top-tier data mining conferences (e.g., WSDM). It analyzes the hybrid single/double-blind architecture, the impact of no rebuttal phase, and the distinct perspectives of academic and industry reviewers. Use this guide to understand the requirements for writing and structuring submissions that maximize acceptance chances.