RLHF causes consistent sentiment drift to neutral in multilingual summarization primarily via KL regularization, which a sentiment-aware KL modification can mitigate.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Multilingual Sentiment Aware Text Summarization A Reinforcement Learning Approach for Consistency Maintenance
RLHF causes consistent sentiment drift to neutral in multilingual summarization primarily via KL regularization, which a sentiment-aware KL modification can mitigate.