Confidence-Reward Driven Preference Optimization for Machine Translation
- Author
- Neural Intelligence Network
- Published
- Sun 09 Feb 2025
- Episode Link
- https://podcasters.spotify.com/pod/show/neuralintelpod/episodes/Confidence-Reward-Driven-Preference-Optimization-for-Machine-Translation-e2u639d
The paper "CRPO: Confidence-Reward Driven Preference Optimization for Machine Translation" introduces a novel approach to improving machine translation (MT) performance by leveraging both reward scores and model confidence for data selection during fine-tuning.
