AI-Research-SKILLs

In the author's own words, from each skill's SKILL.md — showing the first 30 of 98 files.

• autoresearch

  Orchestrates end-to-end autonomous AI research projects using a two-loop architecture. The inner loop runs rapid experiment iterations with clear optimization targets. The outer loop synthesizes results, identifies patterns, and steers research direction. Routes to domain-specific skills for execution, supports continuous agent operation via Claude Code /loop and OpenClaw heartbeat, and produces…

  0-autoresearch-skill/SKILL.md on GitHub (opens in a new tab)

• implementing-llms-litgpt

  Implements and trains LLMs using Lightning AI's LitGPT with 20+ pretrained architectures (Llama, Gemma, Phi, Qwen, Mistral). Use when need clean model implementations, educational understanding of architectures, or production fine-tuning with LoRA/QLoRA. Single-file implementations, no abstraction layers.

  01-model-architecture/litgpt/SKILL.md on GitHub (opens in a new tab)

• mamba-architecture

  State-space model with O(n) complexity vs Transformers' O(n²). 5× faster inference, million-token sequences, no KV cache. Selective SSM with hardware-aware design. Mamba-1 (d_state=16) and Mamba-2 (d_state=128, multi-head). Models 130M-2.8B on HuggingFace.

  01-model-architecture/mamba/SKILL.md on GitHub (opens in a new tab)

• nanogpt

  Educational GPT implementation in ~300 lines. Reproduces GPT-2 (124M) on OpenWebText. Clean, hackable code for learning transformers. By Andrej Karpathy. Perfect for understanding GPT architecture from scratch. Train on Shakespeare (CPU) or OpenWebText (multi-GPU).

  01-model-architecture/nanogpt/SKILL.md on GitHub (opens in a new tab)

• rwkv-architecture

  RNN+Transformer hybrid with O(n) inference. Linear time, infinite context, no KV cache. Train like GPT (parallel), infer like RNN (sequential). Linux Foundation AI project. Production at Windows, Office, NeMo. RWKV-7 (March 2025). Models up to 14B parameters.

  01-model-architecture/rwkv/SKILL.md on GitHub (opens in a new tab)

• distributed-llm-pretraining-torchtitan

  Provides PyTorch-native distributed LLM pretraining using torchtitan with 4D parallelism (FSDP2, TP, PP, CP). Use when pretraining Llama 3.1, DeepSeek V3, or custom models at scale from 8 to 512+ GPUs with Float8, torch.compile, and distributed checkpointing.

  01-model-architecture/torchtitan/SKILL.md on GitHub (opens in a new tab)

• huggingface-tokenizers

  Fast tokenizers optimized for research and production. Rust-based implementation tokenizes 1GB in <20 seconds. Supports BPE, WordPiece, and Unigram algorithms. Train custom vocabularies, track alignments, handle padding/truncation. Integrates seamlessly with transformers. Use when you need high-performance tokenization or custom tokenizer training.

  02-tokenization/huggingface-tokenizers/SKILL.md on GitHub (opens in a new tab)

• sentencepiece

  Language-independent tokenizer treating text as raw Unicode. Supports BPE and Unigram algorithms. Fast (50k sentences/sec), lightweight (6MB memory), deterministic vocabulary. Used by T5, ALBERT, XLNet, mBART. Train on raw text without pre-tokenization. Use when you need multilingual support, CJK languages, or reproducible tokenization.

  02-tokenization/sentencepiece/SKILL.md on GitHub (opens in a new tab)

• axolotl

  Expert guidance for fine-tuning LLMs with Axolotl - YAML configs, 100+ models, LoRA/QLoRA, DPO/KTO/ORPO/GRPO, multimodal support

  03-fine-tuning/axolotl/SKILL.md on GitHub (opens in a new tab)

• llama-factory

  Expert guidance for fine-tuning LLMs with LLaMA-Factory - WebUI no-code, 100+ models, 2/3/4/5/6/8-bit QLoRA, multimodal support

  03-fine-tuning/llama-factory/SKILL.md on GitHub (opens in a new tab)

• peft-fine-tuning

  Parameter-efficient fine-tuning for LLMs using LoRA, QLoRA, and 25+ methods. Use when fine-tuning large models (7B-70B) with limited GPU memory, when you need to train <1% of parameters with minimal accuracy loss, or for multi-adapter serving. HuggingFace's official library integrated with transformers ecosystem.

  03-fine-tuning/peft/SKILL.md on GitHub (opens in a new tab)

• unsloth

  Expert guidance for fast fine-tuning with Unsloth - 2-5x faster training, 50-80% less memory, LoRA/QLoRA optimization

  03-fine-tuning/unsloth/SKILL.md on GitHub (opens in a new tab)

• nnsight-remote-interpretability

  Provides guidance for interpreting and manipulating neural network internals using nnsight with optional NDIF remote execution. Use when needing to run interpretability experiments on massive models (70B+) without local GPU resources, or when working with any PyTorch architecture.

  04-mechanistic-interpretability/nnsight/SKILL.md on GitHub (opens in a new tab)

• pyvene-interventions

  Provides guidance for performing causal interventions on PyTorch models using pyvene's declarative intervention framework. Use when conducting causal tracing, activation patching, interchange intervention training, or testing causal hypotheses about model behavior.

  04-mechanistic-interpretability/pyvene/SKILL.md on GitHub (opens in a new tab)

• sparse-autoencoder-training

  Provides guidance for training and analyzing Sparse Autoencoders (SAEs) using SAELens to decompose neural network activations into interpretable features. Use when discovering interpretable features, analyzing superposition, or studying monosemantic representations in language models.

  04-mechanistic-interpretability/saelens/SKILL.md on GitHub (opens in a new tab)

• transformer-lens-interpretability

  Provides guidance for mechanistic interpretability research using TransformerLens to inspect and manipulate transformer internals via HookPoints and activation caching. Use when reverse-engineering model algorithms, studying attention patterns, or performing activation patching experiments.

  04-mechanistic-interpretability/transformer-lens/SKILL.md on GitHub (opens in a new tab)

• nemo-curator

  GPU-accelerated data curation for LLM training. Supports text/image/video/audio. Features fuzzy deduplication (16× faster), quality filtering (30+ heuristics), semantic deduplication, PII redaction, NSFW detection. Scales across GPUs with RAPIDS. Use for preparing high-quality training datasets, cleaning web data, or deduplicating large corpora.

  05-data-processing/nemo-curator/SKILL.md on GitHub (opens in a new tab)

• ray-data

  Scalable data processing for ML workloads. Streaming execution across CPU/GPU, supports Parquet/CSV/JSON/images. Integrates with Ray Train, PyTorch, TensorFlow. Scales from single machine to 100s of nodes. Use for batch inference, data preprocessing, multi-modal data loading, or distributed ETL pipelines.

  05-data-processing/ray-data/SKILL.md on GitHub (opens in a new tab)

• grpo-rl-training

  Expert guidance for GRPO/RL fine-tuning with TRL for reasoning and task-specific model training

  06-post-training/grpo-rl-training/SKILL.md on GitHub (opens in a new tab)

• miles-rl-training

  Provides guidance for enterprise-grade RL training using miles, a production-ready fork of slime. Use when training large MoE models with FP8/INT4, needing train-inference alignment, or requiring speculative RL for maximum throughput.

  06-post-training/miles/SKILL.md on GitHub (opens in a new tab)

• openrlhf-training

  High-performance RLHF framework with Ray+vLLM acceleration. Use for PPO, GRPO, RLOO, DPO training of large models (7B-70B+). Built on Ray, vLLM, ZeRO-3. 2× faster than DeepSpeedChat with distributed architecture and GPU resource sharing.

  06-post-training/openrlhf/SKILL.md on GitHub (opens in a new tab)

• simpo-training

  Simple Preference Optimization for LLM alignment. Reference-free alternative to DPO with better performance (+6.4 points on AlpacaEval 2.0). No reference model needed, more efficient than DPO. Use for preference alignment when want simpler, faster training than DPO/PPO.

  06-post-training/simpo/SKILL.md on GitHub (opens in a new tab)

• slime-rl-training

  Provides guidance for LLM post-training with RL using slime, a Megatron+SGLang framework. Use when training GLM models, implementing custom data generation workflows, or needing tight Megatron-LM integration for RL scaling.

  06-post-training/slime/SKILL.md on GitHub (opens in a new tab)

• torchforge-rl-training

  Provides guidance for PyTorch-native agentic RL using torchforge, Meta's library separating infra from algorithms. Use when you want clean RL abstractions, easy algorithm experimentation, or scalable training with Monarch and TorchTitan.

  06-post-training/torchforge/SKILL.md on GitHub (opens in a new tab)

• fine-tuning-with-trl

  Fine-tune LLMs using reinforcement learning with TRL - SFT for instruction tuning, DPO for preference alignment, PPO/GRPO for reward optimization, and reward model training. Use when need RLHF, align model with preferences, or train from human feedback. Works with HuggingFace Transformers.

  06-post-training/trl-fine-tuning/SKILL.md on GitHub (opens in a new tab)

• verl-rl-training

  Provides guidance for training LLMs with reinforcement learning using verl (Volcano Engine RL). Use when implementing RLHF, GRPO, PPO, or other RL algorithms for LLM post-training at scale with flexible infrastructure backends.

  06-post-training/verl/SKILL.md on GitHub (opens in a new tab)

• constitutional-ai

  Anthropic's method for training harmless AI through self-improvement. Two-phase approach - supervised learning with self-critique/revision, then RLAIF (RL from AI Feedback). Use for safety alignment, reducing harmful outputs without human labels. Powers Claude's safety system.

  07-safety-alignment/constitutional-ai/SKILL.md on GitHub (opens in a new tab)

• llamaguard

  Meta's 7-8B specialized moderation model for LLM input/output filtering. 6 safety categories - violence/hate, sexual content, weapons, substances, self-harm, criminal planning. 94-95% accuracy. Deploy with vLLM, HuggingFace, Sagemaker. Integrates with NeMo Guardrails.

  07-safety-alignment/llamaguard/SKILL.md on GitHub (opens in a new tab)

• nemo-guardrails

  NVIDIA's runtime safety framework for LLM applications. Features jailbreak detection, input/output validation, fact-checking, hallucination detection, PII filtering, toxicity detection. Uses Colang 2.0 DSL for programmable rails. Production-ready, runs on T4 GPU.

  07-safety-alignment/nemo-guardrails/SKILL.md on GitHub (opens in a new tab)

• prompt-guard

  Meta's 86M prompt injection and jailbreak detector. Filters malicious prompts and third-party data for LLM apps. 99%+ TPR, <1% FPR. Fast (<2ms GPU). Multilingual (8 languages). Deploy with HuggingFace or batch processing for RAG security.

  07-safety-alignment/prompt-guard/SKILL.md on GitHub (opens in a new tab)

No skill matches that filter.

Similar skills

• ui-ux-pro-max-skillby nextlevelbuilder An AI skill that provides design intelligence for building professional UI/UX across multiple platforms. 136k stars · 7 skills
• cavemanby JuliusBrussee 🪨 why use many token when few token do trick. Viral skill + proxy for coding agents that cuts 65% of tokens by talking like a caveman. 113k stars · 20 skills
• Understand-Anythingby Egonex-AI Graphs that teach > graphs that impress. Turn any code into an interactive knowledge graph you can explore, search, and ask questions about… 86.5k stars · 9 skills
• archifyby tt-a1i Turn any idea, plan, or codebase into a beautiful interactive diagram. An agent skill for Claude Code, Codex, and more. 84.1k stars · 2 skills
• oh-my-openagentby code-yeongyu OmO: Just type "mass ulw" keyword with your prompt. Now you are the master of graph engineering. 70.1k stars · 25 skills
• academic-research-skillsby Imbad0202 Academic Research Skills for Claude Code: research → write → review → revise → finalize 52.3k stars · 4 skills