Poster Type: Research Posters
Author: Kartik Shinde (Predii), Tirthankar Ghosal (Oak Ridge National Laboratory (ORNL)), Sreekantam Sai Venkat (Predii), Siddhant Senapati (Predii), Sanjay Das (Oak Ridge National Laboratory (ORNL)), Tilak Kasturi (Predii), Khaled Ibrahim (Lawrence Berkeley National Laboratory (LBNL)), Hugo Hernandez (spinTwo), Kanad Basu (Rensselaer Polytechnic Institute), Sukumar Rangan (Advanced Micro Devices (AMD))
Abstract: Large language models (LLMs) and agentic systems offer promising pathways for automating high performance computing (HPC) operations, yet their real-world deployment is severely constrained by silent operational failures, uncalibrated confidence, and brittle central control. In this work, we chart a new vision for reliable, agentic AI across supercomputing facilities. We explore the architectural paradigm shift from centralized orchestration to decentralized, peer-to-peer agent topologies, illustrating how typed dialogue and bounded knowledge limits ensure defensible, transparent operational guidance. Furthermore, we examine the critical challenge of stance fidelity, demonstrating how verbalized confidence must be audited as dynamic system telemetry rather than taken at face value. Based on these insights, we advocate for a fundamental re-architecting of facility AI: prioritizing honest refusal over raw recall, establishing auditable ground-truth benchmarks, and transitioning AI tools from passive historical log lookups to active, pre-submission job steering.
Best Poster Finalist (BP): no
Poster: PDF
Poster Summary: PDF