What’s Cooking

CoreweaveBlog2w

ART-Optimized Megatron (AOM): 12x higher training throughput

CloudflareBlog2w

Introducing Clef: our open-source decision models, and new RL fine-tuning platform

AppleBlog2w

RLTL;DR: Self-Improvement by Internalizing Self-Generated Feedback

Fireworks AIBlog2w

Reinforcement learning: Why alignment of numerics and MoE routing matter

CoreweaveBlog2w

Getting started with the Agent Reinforcement Trainer (ART)

CoreweaveBlog2w

From Simulation-to-Reality for Robotics

You’re caught up

That’s every post that week in this feed.

21–27 Sep 2026