Goal-Persistent Coding Agents as Scientific Performance Engineers: A Fixed-Radius Nearest-Neighbor Case Study
On the map: LLM agents and tool use → Software engineering coding agents
Abstract (arXiv)
Coding agents can pursue persistent objectives across many tool-use turns, but evidence that general-purpose agents can conduct rigorous scientific performance engineering remains limited. We present a repository-scale case study in which off-the-shelf Codex and Claude Code agents optimize fixed-radius nearest-neighbor (FRNN) search for particle tracking. Starting from a PyTorch-dependent CUDA implementation, the agents follow an executable goal that specifies exact-correctness tests, profiling requirements, and acceptance criteria without prescribing code transformations. In the primary sequential trajectory, they autonomously remove the PyTorch dependency and conduct hypothesis-driven optimization experiments. The resulting standalone C++/CUDA library exactly reproduces the targeted reference result. Its synchronous NumPy interface achieved 1.6-fold speedup over the original GPU-resident PyTorch interface, despite including host transfers. Similar speedups were observed across different GPU architectures and software stacks. An independent optimization rerun followed a different sequence of hypotheses and reached even better performance on the target workload. These results show that goal-persistent coding agents can act as experimental performance engineers, and that executable scientific contracts are needed both to guide and to validate their optimization.
Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Software engineering coding agents, one of 16 directions in LLM agents and tool use. Builds on 6 of its 20 references that are in the map. First posted on arXiv 25 Sep 2026.
Builds on
6 of 20 references.
Closest ancestors · 1
Same direction, newest first
Other references · 5
- 2017PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Spacecited by 4,384
- 2024ReEvo: Large Language Models as Hyper-Heuristics with Reflective Evolutioncited by 266
- 2021Performance of a Geometric Deep Learning Pipeline for HL-LHC Particle Trackingcited by 29
- 2026ScienceFlow: A long-horizon agent for ML research, scientific discovery and beyondcited by 3
- 2026Hypothesis-Driven Autonomous Materials Synthesis with Multimodal LLM Agentscited by 1
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2026KernelBlaster: Continual Cross-Task CUDA Optimization via Memory-Augmented In-Context Reinforcement Learning
- 2026Agentic Kernel Optimization: Generating State-of-the-Art GPU Kernels Without Hand-Written CUDA
- 2026GPUPhysBench: Benchmarking Coding Agents for Correct and Efficient GPU Physics Simulation
- 2026AgentKernelArena: Generalization-Aware Benchmarking of GPU Kernel Optimization Agents
- 2025From Legacy Fortran to Portable Kokkos: An Autonomous Agentic AI Workflow
- 2026CUDA Agent: Large-Scale Agentic RL for High-Performance CUDA Kernel Generation
- 2026PERFOPT-Bench: Evaluating Coding Agents on Software Performance Optimization
- 2026Ave: Guiding Agentic GPU Optimization Using Data-Flow Invariants