How I Study AI - Learn AI Papers & Lectures the Easy Way

CUDA Agent: Large-Scale Agentic RL for High-Performance CUDA Kernel Generation

Intermediate

Weinan Dai, Hanlin Wu et al.Feb 27arXiv

CUDA Agent is a training system that teaches an AI to write super-fast GPU code (CUDA kernels) by practicing, testing, and getting rewards for correct and speedy results.

#CUDA kernel generation#agentic reinforcement learning#PPO actor-critic

Dr. Kernel: Reinforcement Learning Done Right for Triton Kernel Generations

Intermediate

Wei Liu, Jiawei Xu et al.Feb 5arXiv

This paper teaches a language model to write fast GPU kernels (tiny speed programs) in Triton using reinforcement learning that really cares about meaningful speed, not just being correct.

#Triton kernels#Reinforcement learning#Policy gradient

PACEvolve: Enabling Long-Horizon Progress-Aware Consistent Evolution

Intermediate

Minghao Yan, Bo Peng et al.Jan 15arXiv

PACEvolve is a new recipe that helps AI agents improve their ideas step by step over long periods without getting stuck.

#evolutionary search#LLM agents#context management

Papers3

CUDA Agent: Large-Scale Agentic RL for High-Performance CUDA Kernel Generation

Dr. Kernel: Reinforcement Learning Done Right for Triton Kernel Generations

PACEvolve: Enabling Long-Horizon Progress-Aware Consistent Evolution