Calculates the number of disorder pairs (inversions) in a list where i < j and pi > pj, optimized for large datasets.

Designs custom Pokémon evolution lines or converts video game characters into Pokémon profiles, including types, abilities, stats, lore, appearance, and portmanteau names.

Rewords and paraphrases vocational training performance criteria and adds concrete workplace examples to strengthen them.

Interview the user to flesh out a fictional character's personality and psychological profile, focusing on internal traits, motivations, and behaviors rather than just logistical details.

Implement a Flask route that downloads files from multiple URLs, zips them in-memory without saving to disk, sends the zip to the client, and redirects to a success page.

Answer the IELTS Speaking Part 2 topic about a happy childhood memory by strictly adhering to the required structure: describing the event, time and place, companions, and the reason for happiness.

Creates structured job profiles including company overview, role requirements, responsibilities, skills, location, and salary range.

Создание пользовательского слоя nn.Module, веса которого являются обучаемыми параметрами (torch.nn.Parameter) и обновляются через loss.backward() и optimizer.step(), вместо статического вычисления на входе.

Rewrites and optimizes academic research titles in the field of nutrition to be search-friendly, concise, and easy to understand, adopting the persona of a nutrition researcher or professor.

Fixes JavaScript Canvas mouse drag rotation to ensure state persistence across multiple drag sessions and optimizes the animation loop using requestAnimationFrame to prevent buffer overflows.

Conducts interviews to develop a character's psychological profile by asking questions about personality, motivations, emotional responses, and interpersonal dynamics.

Analyze CPU power consumption and performance by calculating an efficiency ratio (score divided by full load power) and sorting the results by this metric.

Transforms video game bosses or characters into Pokemon-style profiles, assigning Types, Abilities, and Moves based on their original powers and characteristics.

Rewords or paraphrases performance criteria for training contexts and adds concrete, observable workplace examples to strengthen assessment context.

Provides concise instructions for unlocking LUKS drives, mounting Btrfs RAID 1 in degraded mode, and converting to a single-drive profile within a Debian initramfs prompt.

Define and structure the comprehensive requirements for a government contract management process, covering roles, performance metrics, risk management, change control, subcontractor oversight, compliance, communication, and documentation.

Analyzes job positions to define scope, target companies, salary, and candidate profiles, while generating structured job descriptions and comprehensive interview guides for both candidates and interviewers.

Analyzes a user's preferred consulting role from Lippit and Lippit's Guidance model by defining all roles, summarizing user values/interests, and explaining the alignment between the chosen role and the user's profile.

Analyzes a psychological profile defined by a PVEL code and function stack (Dominant to Inferior), describing the resulting mindset dynamics and assigning corresponding MBTI and Enneagram types.

Создание класса nn.Module для понижения размерности (например, на основе QR-разложения), который является обучаемым (использует torch.nn.Parameter и обновляется через optimizer.step).

Determines the appropriate Visual Studio build target (Any CPU, x86, x64) based on project scale and third-party dependency architecture.

Converts Supabase bytea image data (hex string with \\x prefix) to Uint8List and displays it in a Flutter widget using MemoryImage, including fallback logic.

Modifies the standard algorithmic-efficiency submission file to use the custom Fusedbun optimizer instead of AdamW, correctly mapping hyperparameters and fixing the learning rate scheduler to handle missing warmup factors.

Translate Japanese song lyrics into English and assess their difficulty across three dimensions: translation complexity, cultural/poetic understanding, and singing performance requirements.

Develops parallel Java code for a given problem and benchmarks performance by varying thread counts from 1 to 16, repeating experiments 5 times, and reporting individual run times and averages.

Optimizes memory consumption during PyTorch model training by implementing mixed precision training, gradient accumulation, and efficient data loading strategies to fit within hardware constraints.

Write engaging blog posts optimized for search engines, maintaining a warm and friendly tone while emphasizing pro-business and pro-American values.

Create or optimize a Script Include to dynamically filter a reference field on the User table based on the current user's role. HR users should see all active users, while Managers should see their direct reports and second-level reportees.

Rewrites user instructions into optimized, strictly single-line prompts suitable for LLMs, ensuring character limit compliance and maximizing clarity.

Simulates a side-scrolling beat 'em up RPG where the user navigates levels and fights enemies using stats (HP, Power, Defense, Speed). Tracks health, calculates combat outcomes based on move success chances, and presents choices with hidden outcomes and varying favorability.

A Python program to detect anomalies in videos using the VideoMAEForPreTraining model. It processes videos by dividing them into 16-frame clips, extracts embeddings using an unmasked boolean mask, and compares them against a normal behavior profile using Mean Squared Error (MSE).

Automates the process of training multiple neural network instances with varying configurations (sizes, layers, dimensions) sequentially and compares their performance metrics at the end.

Configures an autonomous AI agent to execute tasks efficiently using a strict JSON command interface, comprehensive tool usage, and memory management. Handles short-term memory limits and random shutdowns via persistent file storage and structured reasoning.

Facilitates a text-based, side-scrolling beat 'em up RPG. Tracks stats (HP, Power, Defense, Speed), manages combat encounters with calculated outcomes, and presents narrative choices with hidden results.

Develop a BERT-based pipeline to classify speakers (agent vs. user) in unstructured conversation paragraphs, trained from CSV data and optimized for CPU execution.

Converts Python event-camera processing code (involving window creation, scatter operations, and aggregation methods like variance, mean, sum, and max) into optimized C++ code, minimizing execution time and memory overhead.

Генерация надежного кода C++ (DLL или клиент драйвера) для патчинга памяти процесса с поддержкой RAII, Unicode, асинхронного выполнения и целевого патчинга N-го вхождения.

A structured framework for managing conversations using session-based memory, sentiment analysis, and adaptive response generation to ensure empathetic and coherent engagement.

Evaluate the risk profile of IT or organizational assets by determining threat values, vulnerability levels, likelihood of occurrence, risk scores, and appropriate treatments, strictly adhering to defined output value sets.

Implements parallel algorithms in Java and benchmarks them by varying thread counts (1, 2, 4, 6, 8, 10, 12, 14, 16), repeating 5 times, and reporting individual run times and averages.

Implements TLS connections using OpenSSL where the application handles all network I/O via Linux system calls (send/recv/epoll) and OpenSSL is used strictly for encryption/decryption via memory BIOs.

Modifies PyTorch data preparation scripts for RNN/LSTM models to limit the dataset size by dividing it into chunks controlled by a hyperparameter, ensuring the first dimension of input/target tensors fits memory constraints.

Provides a complete end-to-end workflow for text classification using a PyTorch Transformer model. It includes automatic vocabulary generation from raw text, a custom tokenizer implementation, data padding, model training on CPU, and visualization of loss and accuracy metrics.

Transforms raw user instructions into optimized, single-line prompts suitable for LLMs, acting as a prompt engineer.

Implement comprehensive logging, checkpointing, and visualization for a Reinforcement Learning training loop, tracking rewards, losses, actions, states, entropy, and performance metrics using CSV and log files.

Develops a modular Unity system for ramp interaction consisting of a player-side RampDetection script and a ramp-side Ramp configuration script. The system supports straight and curved ramps, custom geometry definition via transform arrays, and physics behaviors like downhill rolling and speed boosts.

Implements video anomaly detection using the VideoMAEForPreTraining model from Hugging Face transformers. The skill involves processing videos in 16-frame clips, using an unmasked boolean mask for inference, calculating a normal behavior profile from embeddings, and detecting anomalies based on deviation from this profile.

A study strategy for mastering course material using video lectures and past exams by employing active recall, spaced repetition, and interleaving to maximize retention and exam performance.

Genera paquetes de contenido integral para YouTube en español, incluyendo títulos SEO, descripciones, sugerencias de miniaturas, prompts de IA y palabras clave para Google Ads, optimizando para la visibilidad y el compromiso.

编写Python脚本以定时监控Windows主机的CPU和内存使用率，并将包含时间戳、CPU占用率和内存占用率的数据记录到日志文件中。