ROAM freezes specialist models and uses LLM priors plus online evidence in a 5-D semantic latent space to cut major-shift MAE by over 20% with under 0.02 ms overhead.
Bayesian hierarchical graph neural networks with uncertainty feedback for trust- worthy fault diagnosis of industrial processes,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Open-Ended Scenario Reasoning for Specialist Model Adaptation
ROAM freezes specialist models and uses LLM priors plus online evidence in a 5-D semantic latent space to cut major-shift MAE by over 20% with under 0.02 ms overhead.