GPUAlert wraps training commands to deliver classified failure reports, durable logs, and preserved exit codes with 0.997 macro-F1 on reproduced classes and 3 ms overhead.
Hellerstein, and Aditya G
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.SE 2representative citing papers
The submission's abstract claims a CPG-based ML smell detector with 88.14% recall, but the full text implements an AST-only DSL detector with 88.89% recall, and no CPG experiment appears.
citing papers explorer
-
GPUAlert: A Zero-Instrumentation Process-Boundary Monitor for Diagnosing GPU Training-Job Failures
GPUAlert wraps training commands to deliver classified failure reports, durable logs, and preserved exit codes with 0.997 macro-F1 on reproduced classes and 3 ms overhead.
-
SpecDetect4ML: Detecting Non-Local ML Code Smells with Code Property Graphs
The submission's abstract claims a CPG-based ML smell detector with 88.14% recall, but the full text implements an AST-only DSL detector with 88.89% recall, and no CPG experiment appears.