Summary
Scientific computing often involves complicated steps like setting up environments, requesting resources, and fixing errors. The authors developed ASCEND, an AI agent that runs on a researcher’s laptop to remotely manage scientific jobs on computing clusters and GPU workstations, following local policies without needing special central services. The agent successfully handled tasks like diagnosing and fixing job failures, reproducing published scientific results with high accuracy, speeding up long-running computations, and detecting bugs in existing codes. This approach helps scientists by automating routine and error-prone parts of running complex computations.
What this means in practice
- •For high-performance computing operators: Automatically run, monitor, and recover scientific computational jobs across HPC clusters with minimal user intervention and without central infrastructure.
- •For data engineering teams: Speed up large-scale geophysical or scientific solvers by safely automating parallelization under strict correctness requirements.
Abstract
Traditional scientific computing requires researchers to translate computational intent into environment configuration, resource requests, and executable jobs, then diagnose failures from scheduler state and application logs. We present ASCEND (Autonomous Scientific Computing Engine and Novel Discovery), an AI-powered agent interface that runs the agent on the researcher's own laptop, reaching Slurm-managed clusters and a GPU workstation over a multiplexed authenticated connection, with site-specific execution policies checked by locally executed tools; the language model is hosted remotely and holds no credentials. No facility-scale service is required: an account on each resource is sufficient, and the public installer lets users link additional Slurm clusters or workstations of their own. We report four recorded cases: (1) the agent closed a failure-recovery loop on a planted tensor-device fault, submitting, diagnosing, repairing and resubmitting with job-level artifacts preserved; (2) it reproduced the published evaluation of a weather-forecasting model from the author's released forecasts, agreeing with the published curves to 2.1% (z500) and 2.4% (t850) while identifying a unit discrepancy in the paper's prose and an initialization-field discrepancy in its released data; (3) it parallelized a released 12,693-line geophysical solver under a bit-for-bit identity requirement, reducing wall-clock runtime from about twelve hours to about two; (4) that requirement exposed two instances of undefined behaviour in the published solver, both repaired and reported upstream. Separately, a pre-specified evaluation of the policy layer found the deployed validator rejected 29 of 30 constructed violations and held the remaining one for approval, while denying 3 of 14 legitimate requests. Autonomy was exercised under author supervision; an end-to-end recovery benchmark remains outstanding.