A GRU plus attention plus feedforward classifier outperforms transformer baselines on Azure telemetry fault prediction in the reported metrics, but without code, error bars, or train/test details.
Structured preference modeling for reinforcement learning-based fine-tuning of large models,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.DC 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Time-Series Learning for Proactive Fault Prediction in Distributed Systems with Deep Neural Structures
A GRU plus attention plus feedforward classifier outperforms transformer baselines on Azure telemetry fault prediction in the reported metrics, but without code, error bars, or train/test details.