Answers are grounded in this paper and cite their sources.
Large language models continue to advance, with an increasing number of domain-specific variants tailored for specialised tasks. However, these models often lack transparency and explainability, can be costly to fine-tune, require substantial prompt engineering, yield inconsistent results across domains, and impose a significant adverse environmental impact due to their high computational demands. To address these challenges, we propose the Bayesian network pre-trained language model fusion (BNPL) framework, which integrates predictions from three pre-trained language models, including FinBERT, RoBERTa, and BERTweet, through a probabilistic mechanism for sentiment analysis. BNPL performs late fusion by modelling the sentiment predictions from these models as probabilistic nodes within a Bayesian network. Evaluated across three human-annotated financial corpora with distinct linguistic characteristics, BNPL improves classification performance over individual PLMs on the combined test set and provides decision-level transparency by quantifying model dependencies, with low computational overhead at the fusion stage.
No takes yet. Share an insight, caveat, or question.
Thiruvady et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: