KernelAudit: Contract-Aware Auditing of Validators for LLM-Generated GPU Kernels
cs.SE
Submitted: 2026-11-11
Updated: 2026-11-11
Code: https://github.com/ZIYUANLIstar/KernelAudit
Terminology
Sources
- Concrete Problems in AI Safety
- MusaCoder: Native GPU Kernel Generation with Full-Stack Training on Moore Threads GPU
- CUDA Agent: Large-Scale Agentic RL for High-Performance CUDA Kernel Generation
- Equivalence Checking of ML GPU Kernels
- Towards Robust Agentic CUDA Kernel Benchmarking, Verification, and Optimization
- Correct but Slow: An Empirical Study of the GPU Kernel Evaluation Gap in Modern Domain-Specific Languages
- Dr. Kernel: Reinforcement Learning Done Right for Triton Kernel Generations
- KernelBench: Can LLMs Write Efficient GPU Kernels?
- Characterizing Overconfident Failure in LLM-Based Code Generation
- Before the Model Learns the Bug:Fuzzing RLVR Verifiers
- The Correctness Illusion in LLM-Generated GPU Kernels
- A Contract-Grade Verifier for LLM-Generated GPU Kernels, and a Native Blackwell Backward for the Gated-Linear-Recurrence Family
- Harness Engineering for LLM-Driven GPU Kernel Generation
- Kernel Contracts: A Specification Language for ML Kernel Correctness Across Heterogeneous Silicon
- KernelBenchX: A Comprehensive Benchmark for Evaluating LLM-Generated GPU Kernels
- Geak: Introducing Triton Kernel AI Agent & Evaluation Benchmarks
- Are LLM-Generated GPU Kernels Production-Ready? A Trace-Driven Benchmark and Optimization Agent
- KernelBench-Verified: Do LLM-Generated Kernels Actually Beat PyTorch?
Related papers
- ArkEval: Benchmarking and Evaluating Automated CodeRepair for ArkTS
- TestPrism: Rethinking Test Evaluation Beyond a Single Reference
- Mining Workflow Graphs for Black-Box Boundary Testing of Conversational LLM Agents
- Falsification-Based Verification of LLM-Generated Optimization Models: Sound Test Batteries and Their Detection Limits
- GitSkills: A Dataset of Agent Skills on GitHub
- For Coding Agents, the Repository Matters