Projects Recently Starred by Experts

Starred 20 hours ago by

Pawel Garbacki(Cofounder of Fireworks AI). Also starred by

Shizhe Diao(Author of LMFlow; Research Scientist at NVIDIA) and

Elvis Saravia(Founder of DAIR.AI).

OpenResearcher by TIGER-AI-Lab

14.1%

482

Agentic LLM for long-horizon deep research trajectory synthesis

Empowers complex research tasks by synthesizing multi-turn, web-browsing trajectories.
Leverages a 30B-A3B model trained on a large-scale, open-sourced deep research dataset.
Features a scalable, low-cost retriever and an open evaluation framework for reproducible research.
Outperforms leading models like GPT-4 and Gemini 2.5 Pro on key deep research benchmarks.

Created 1 month ago

Updated 23 hours ago

Starred 20 hours ago by

Theo Browne(Founder of Ping.gg). Also starred by

Vincent Weisser(Cofounder of Prime Intellect) and

Calvin French-Owen(Cofounder of Segment).

parameter-golf by openai

178.6%

Extreme parameter-constrained LM training challenge

Train LMs under a strict 16MB artifact size limit.
Focus on novel architectures, compression, and fast training (<10 min).
Supports local MLX development and cloud GPU scaling.
Competition for frontier AI research and talent identification.

Created 1 month ago

Updated 19 hours ago

Starred 21 hours ago by

Sebastian Raschka(Author of "Build a Large Language Model (From Scratch)"). Also starred by

Pawel Garbacki(Cofounder of Fireworks AI),

Wes McKinney(Author of Pandas),

Jordan Burgess(Cofounder of Humanloop), and

14 more.

chainlit by Chainlit

0.4%

12k

Framework for rapid conversational AI app development

Python framework to build chatbot UIs.
Integrates with LangChain, LlamaIndex, OpenAI, etc.
Supports streaming, auth, and rich media.

Created 3 years ago

Updated 2 days ago

Starred 21 hours ago by

Gabriel Almeida(Cofounder of Langflow).

HyperAgents by facebookresearch

880

Self-improving agents for optimizing computable tasks

Develops self-referential agents capable of optimizing any computable task through iterative self-improvement.
Leverages foundation models via API keys (OpenAI, Anthropic, Gemini) for agentic reasoning and code generation.
Implements a meta-agent and task-agent architecture within a generative loop for autonomous optimization.
Executes model-generated code, necessitating awareness of potential safety risks and model limitations.

Created 1 week ago

Updated 3 days ago

Starred 1 day ago by

Toran Bruce Richards(Founder of AutoGPT).

skills by slavingia

N/A

AI-powered Claude Code plugin for entrepreneurial guidance

Guides users through business ideation, validation, and customer acquisition.
Provides AI-driven skills for pricing, marketing, and sustainable growth.
Implements a structured framework based on "The Minimalist Entrepreneur" book.
Supports decision-making for scaling, culture, and business reviews.

Created 2 days ago

Updated 22 hours ago

Starred 1 day ago by

Travis Fischer(Founder of Agentic). Also starred by

Thomas Dohmke(Former CEO of GitHub) and

Dan Guido(Cofounder of Trail of Bits).

last30days-skill by mvanhorn

56.5%

Real-time AI research and prompt generation assistant

Researches Reddit and X for the last 30 days, synthesizing community insights into actionable prompts.
Generates copy-paste-ready prompts for diverse AI tools, adapting to current community best practices.
Leverages OpenAI and xAI APIs for comprehensive, real-time data acquisition and analysis.
Keeps users current with AI advancements by analyzing trending topics and viral content.

Created 2 months ago

Updated 2 days ago

Starred 1 day ago by

Chaoyu Yang(Founder of Bento). Also starred by

Junyang Lin(Core Maintainer at Alibaba Qwen),

Taranjeet Singh(Cofounder of Mem0),

Ficus Kirkpatrick(Cofounder of /dev/agents; Ex-VP AR/VR at Meta), and

14 more.

superpowers by obra

14.2%

112k

AI assistant superpowers via a comprehensive skills library

Enhances AI agents with systematic debugging, TDD, and collaborative workflows.
Integrates slash commands for planning, execution, and interactive design refinement.
Facilitates parallel agent dispatching and semantic search of past conversations.
Promotes skill creation and contribution using TDD and evidence-based practices.

Created 5 months ago

Updated 18 hours ago

Starred 1 day ago by

Pawel Garbacki(Cofounder of Fireworks AI). Also starred by

Wing Lian(Founder of Axolotl AI),

Stas Bekman(Author of "Machine Learning Engineering Open Book"; Research Engineer at Snowflake), and

Luis Capelo(Cofounder of Lightning AI).

autokernel by RightNow-AI

15.8%

825

Autonomous GPU kernel optimization for PyTorch

Optimizes PyTorch models by iteratively refining Triton or CUDA C++ GPU kernels.
Supports 9 core DL kernel types: matmul, attention, MLPs, and more.
Uses agent-based autoresearch, Amdahl's Law, and rigorous correctness verification.
Exports optimized kernels to HuggingFace Hub for seamless integration.

Created 2 weeks ago

Updated 1 week ago

Starred 1 day ago by

Ross Wightman(Author of timm; CV at Hugging Face).

TalkingHead by met4citizen

1.4%

JavaScript class for real-time lip-sync using 3D avatars

Supports Ready Player Me avatars (GLB) and Mixamo animations (FBX).
Integrates with Google Cloud TTS, ElevenLabs, and Microsoft Azure Speech SDK.
Built-in lip-sync for English, Finnish, and Lithuanian; extensible to 100+ languages.
Includes dynamic bones, custom poses, moods, gestures, and emojis.

Created 2 years ago

Updated 4 weeks ago

Starred 1 day ago by

Jonathan Ragan-Kelley(Professor at MIT). Also starred by

Jeff Hammerbacher(Cofounder of Cloudera) and

Jane Manchun Wong(Security Researcher).

hypura by t8

484

LLM inference scheduler for Apple Silicon, optimizing large models across GPU, RAM, and NVMe

Run models exceeding Mac memory by intelligently scheduling tensors across GPU, RAM, and NVMe storage tiers.
Leverages NVMe for on-demand expert/FFN streaming with high cache hit rates, drastically reducing I/O.
Optimizes tensor placement using model architecture insights and hardware profiling for zero overhead on in-memory models.
Provides an Ollama-compatible server for seamless integration with existing LLM tools and local inference workflows.

Created 1 week ago

Updated 4 days ago

...