Papers6

#Reinforcement learning

CADEvolve: Creating Realistic CAD via Program Evolution

Maksim Elistratov, Marina Barannikov et al.Feb 18arXiv

AI models that make CAD designs used to learn mostly from simple “draw-then-extrude” examples, so they struggled with real, complex parts.

#CAD#CadQuery#Image2CAD

LatentChem: From Textual CoT to Latent Thinking in Chemical Reasoning

Intermediate

Xinwu Ye, Yicheng Mao et al.Feb 6arXiv

LatentChem lets AI do chemistry thinking quietly inside continuous vectors instead of writing long step-by-step sentences.

#Latent reasoning#Chain-of-Thought#Chemical LLM

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

Towards Automated Kernel Generation in the Era of LLMs

Intermediate

Yang Yu, Peiyu Zang et al.Jan 22arXiv

AI programs called LLMs can now help write the tiny, super-fast pieces of code (kernels) that make GPUs run AI models efficiently.

#LLM-driven kernel generation#GPU kernels#CUDA

ChartVerse: Scaling Chart Reasoning via Reliable Programmatic Synthesis from Scratch

Intermediate

Zheng Liu, Honglin Lin et al.Jan 20arXiv

ChartVerse is a new way to make lots of tricky, realistic charts and perfectly checked questions so AI can learn to read charts better.

#Chart reasoning#Vision-Language Models#Rollout Posterior Entropy

OpenTinker: Separating Concerns in Agentic Reinforcement Learning

Intermediate

Siqi Zhu, Jiaxuan YouJan 12arXiv

OpenTinker is an open-source system that makes training AI agents with reinforcement learning simple, modular, and reusable.

#Reinforcement learning#LLM agents#Agent–environment interaction