An experimental PlayStation 5 emulator for Windows, Linux and macOS.

Apple Silicon GPU acceleration for scikit-learn — drop-in Metal backend via zero-copy MPS and custom compute kernels

RAG Telegram bot answering questions about ENSIA Impact community content

IO-aware batched K-Means for Apple Silicon, ported from Flash-KMeans (Triton/CUDA) to pure MLX. Up to 94x faster than sklearn.

Train transformer language models with reinforcement learning.

AI Agent Framework, the Pydantic way

A generalist autonomous research agent — runs experiments, researches, and iteratively optimizes, autonomously.

Self-hosted AI workspace.

Turn any codebase, with its docs, SQL schemas, configs, and PDFs, into a queryable knowledge graph. A /graphify skill for Claude Code, Cursor, Codex, and Gemini CLI: local deterministic AST parsing…

User-friendly AI Interface (Supports Ollama, OpenAI API, ...)

Conditional Memory via Scalable Lookup: A New Axis of Sparsity for Large Language Models

Tracking when and how attention heads specialize during transformer training - A mechanistic interpretability research project with complete implementation, probing pipeline, and analysis tools.

let types trickle from runtime to compile time

Mamba SSM architecture

A novel RNN architecture that solves the forgetting problem via independent gating and additive accumulation.

Paper2Code: Automating Code Generation from Scientific Papers in Machine Learning

Hierarchical Reasoning Model Official Release

🎒 Token-Oriented Object Notation (TOON) – compact, human-readable serialization of JSON data for LLM prompts. TypeScript SDK, CLI, benchmarks.

An Open Phone Agent Model & Framework. Unlocking the AI Phone for Everyone

Minimal CLI coding agent by Mistral

DziriBERT: a Pre-trained Language Model for the Algerian Dialect

Confident Adaptive Selective Momentum Optimizer - Noise-robust PyTorch optimizer