SCOUT: Symmetric Consensus Outlier Detection for Failure Localization in LLM Pre-Training
In LLM pre-training, synchronization propagates rank-local stalls, slowdowns, and numerical errors into job-wide symptoms, obscuring their origin. Existing diagnosis often relies on in-process monitors that cannot report after the trainer blocks or terminates, or on post-mortem logs that preserve only synchronized symptoms; offline health tests lose the workload and operating conditions that triggered the failure. We present SCOUT, a unified runtime failure-localization framework built on one de
Record details
Published: 11 August 2026
Source: arXiv cs.LG
Category: Research
Topics: Jobs & economy · Healthcare
Retrieved: 12 August 2026
Related evidence
These records share source-supplied organisations, an exact publisher byline, automatic topics or regions. The reason is shown on every link; related does not mean supporting, agreeing with or verifying this record.
Automated Features, Algorithms, and Technologies of Electronic Early Warning/Track-and-Trigger Systems: Systematic Review
JMIR (Journal of Medical Internet Research) · 10 August 2026
Deep learning in precision phytopathology: a comprehensive survey of CNN architectures for disease detection and severity quantification
Artificial Intelligence Review · 8 August 2026
Tracing the Heart: An Evidence-Linked Pipeline for Heart-Failure Feature Engineering
arXiv cs.AI · 6 August 2026
Hidden Underbelly of the Silicon Valley: Algorithmic Exploitation and Health in Data Work Value Chains
arXiv cs.CY · 6 August 2026
AI governance at work: Threat appraisal, work experience, and satisfaction with job, health, and financial situation
Technology in Society · 3 August 2026
DualDiT: A Conditional Dual-Output Diffusion Transformer for Joint OCT Image and Segmentation Mask Generation
arXiv cs.AI · 31 July 2026
How to cite this record
ethics.ai (11 August 2026), “SCOUT: Symmetric Consensus Outlier Detection for Failure Localization in LLM Pre-Training,” evidence record 18694, https://ethics.ai/record/18694 (originally published by arXiv cs.LG).
Use and limitations
This page is a stable index and citation surface for a source record. ethics.ai did not author the underlying report and has not independently verified every claim. Automatic topics may be imperfect. For consequential use, quote and cite the original publisher.