CUDA-L2: Surpassing cuBLAS Performance for Matrix Multiplication through Reinforcement Learning
cs.LG, cs.AI
Submitted: 2025-12-02
Updated: 2026-11-11
Code: https://github.com/ornith-ai/CUDA-L2
Terminology
Sources
- GPT-4 Technical Report
- Kevin: Multi-Turn RL for Generating CUDA Kernels
- Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities
- DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
- CUDA-L1: Improving CUDA Optimization via Contrastive Reinforcement Learning
- DeepSeek-V3 Technical Report
- KernelBench: Can LLMs Write Efficient GPU Kernels?
- Instruction Tuning with GPT-4
- DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
- Instruction Tuning for Large Language Models: A Survey
- Gemini: A Family of Highly Capable Multimodal Models
- SwizzlePerf: Hardware-Aware LLMs for GPU Kernel Performance Optimization
- Qwen3 Technical Report
Related papers
- A Physics-Regulated Neural Framework for Learning 3D Grain Growth Dynamics
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis