Models/Systems

INTELLECT-2 - Prime Intellect - 05/12/25

• Organization: Prime Intellect
• Paper:INTELLECT-2: A Reasoning Model Trained Through Globally Decentralized Reinforcement Learning
• Blog: INTELLECT-2 Release: The First 32B Parameter Model Trained Through Globally Distributed Reinforcement Learning
• Hugginface Model Card: INTELLECT-2

Releasing INTELLECT-2: We’re open-sourcing the first 32B parameter model trained via globally distributed reinforcement learning:

• Detailed Technical Report
• INTELLECT-2 model checkpointhttps://t.co/iHDDHRyKN2

— Prime Intellect (@PrimeIntellect) May 12, 2025

SWE-1 - Windsurf - 05/15/25

• Announcement: SWE-1: Our First Frontier Models

The Open Molecules 2025 (OMol25) Dataset, Evaluations, and Models

• Announcement: Sharing new breakthroughs and artifacts supporting molecular property prediction, language processing, and neuroscience
• Hugging Face model card:
  • facebook/OMol25
  • facebook/UMA
• Hugging Face collections: FAIR Chemistry
• Paper:
  • UMA: A Family of Universal Models for Atoms
  • The Open Molecules 2025 (OMol25) Dataset, Evaluations, and Models

Announcing the newest releases from Meta FAIR. We’re releasing new groundbreaking models, benchmarks, and datasets that will transform the way researchers approach molecular property prediction, language processing, and neuroscience.

1️⃣ Open Molecules 2025 (OMol25): A dataset… pic.twitter.com/PAmnNgTVnB

— AI at Meta (@AIatMeta) May 14, 2025

Papers

Current

• Qwen3 Technical Report - 05/15/25
• INTELLECT-2: A Reasoning Model Trained Through Globally Decentralized Reinforcement Learning - 05/12/25
• Continuous Thought Machines
  • Announcement: Introducing Continuous Thought Machines
  • Code: continuous-thought-machines
• Continuous Visual Autoregressive Generation via Score Maximization
  • Code: Continuous Visual Autoregressive Generation via Score Maximization
• BLIP3-o: A Family of Fully Open Unified Multimodal Models-Architecture, Training and Dataset
• HealthBench: Evaluating Large Language Models Towards Improved Human Health

Older

• Sheaf Neural Networks

Repositories

• deer-flow: Open source deep research framework
• LMEval: Large Model Evaluation Framework
  • Announcing LMEval: An Open Source Framework for Cross-Model Evaluation
• astral-sh/ty
  • Announcement: ty - extremely fast type checker

Articles

• What Is ChatGPT Doing … and Why Does It Work?
• Vision Language Models (Better, Faster, Stronger) - 05/12/25
• Building, launching, and scaling ChatGPT Images - 05/13/25
• Build and train a recommender system in 10 minutes using Keras and JAX- 05/13/25
• The Misguided Quest for Mechanistic AI Interpretability - 05/15/25
• Introducing HealthBench - OpenAI - 05/12/25
• Building software on top of Large Language Models
• The Transformers Library: standardizing model definitions - 05/15/25

Courses

• Hugging Face MCP Course

Products

• Introducing Codex

Other

• Generative AI Handbook - by William Brown
• AI Engineering Best Practices

Spoke with 100 startup CTOs today who recognize that embracing AI engineering will soon distinguish winners from losers.

These 15 slides will give you the 10% of game you need to be an effective AI engineer 🧵 pic.twitter.com/hdyoblVSdW

— Nick (@nickbaumann_) May 15, 2025