Skill UI

Activation-aware weight quantization that compresses 7B-70B models to 4-bit with minimal accuracy loss, delivering ~3x inference speedups and memory reductions for constrained GPU deployments while integrating with vLLM and Marlin kernels.

Enables Claude to configure neural network architectures through the neural-network-builder plugin, creating or modifying models and defining layers, activation functions, and training parameters whenever network design or tuning is requested.

A comprehensive guide detailing the installation, setup, and authentication process for Cursor IDE across macOS, Linux, and Windows. It covers account sign-in, plan activation, and the seamless migration of settings and extensions from VS Code, ensuring a fully configured development environment.

This runbook provides comprehensive, step-by-step procedures for managing critical incidents involving Evernote integrations. It covers diagnosing API outages, resolving persistent rate limits, handling authentication failures, and recovering from data synchronization issues. Use this guide during production troubleshooting or service outages to ensure rapid and systematic incident response, including circuit breaker activation and graceful degradation strategies.

Guides researchers through nnsight’s tracing and activation APIs so you can inspect or modify PyTorch internals locally or via NDIF-powered remote runs on 70B+ foundation models without local GPUs.

A comprehensive guide to the Obsidian event system, detailing how to handle various lifecycle and state changes within plugins. Learn to register listeners for vault events (file creation, modification, deletion), workspace events (layout changes, view activation), and metadata cache updates. Essential for building highly reactive and deeply integrated Obsidian plugins.

Expert playbook for accelerating user onboarding, smoothing activation flows, leveraging checklists, empty states, and multi-channel nudges to turn signups into retained customers quickly.

Guides users through plantuml diagram generator tasks, covering activation triggers, best practices, and production-ready code for visual content work.

Pyvene is a declarative framework designed for performing causal interventions on PyTorch neural networks. It allows researchers to conduct advanced experiments such as activation patching, causal tracing (ROME-style), and interchange intervention training (IIT). Use this library when you need to test causal hypotheses about model behavior, deeply interpret model components, or ensure reproducibility in advanced AI research.

Guides optimizing account signup, registration, and trial activation flows to reduce friction, improve completion rates, and drive activation with principles covering field prioritization, trust signals, multi-step best practices, and mobile considerations.

Provides a comprehensive guide for crafting and maintaining Claude Code skills, covering auto-activation hooks, skill-rules configuration, guardrails, domain skills, trigger testing, and Anthropic-aligned best practices such as the 500-line rule and progressive disclosure.

SAELens provides a framework for training and analyzing Sparse Autoencoders (SAEs). SAEs decompose the dense, often polysemantic activations of large language models into sparse, monosemantic features. Use this when you need to discover the discrete, interpretable concepts a model has learned, study feature superposition, or analyze specific safety-relevant behaviors (like bias or deception) within deep neural networks.