Large language models (LLMs) possess strong multilingual capabilities, and combining Reinforcement Learning from Human Feedback (RLHF) with translation tasks has shown great potential. However, we observe that this paradigm performs unexpectedly poorly when applied to colloquial subtitle translation tasks. In this work, we investigate this issue and find that the offline reward model (RM) gradually diverges from the online LLM due to distributional shift, ultimately leading to undesirable training outcomes. To address this, we propose RIVAL, an adversarial training framework that formulates the process as a min–max game between the RM and the LLM. RIVAL iteratively updates the both models, with the RM trained to distinguish strong from weak translations (qualitative preference reward), and the LLM trained to enhance its translation for closing this gap. To stabilize training and improve generalizability, we also incorporate quantitative preference reward (e.g., BLEU) into the RM, enabling reference-free quality modeling aligned with human evaluation. Through extensive experiments, we demonstrate that the proposed training framework significantly improves upon translation baselines.

Visually rich documents (VRDs) are ubiquitous in daily business and life. Examples are purchase receipts, insurance policy documents, custom declaration forms and so on. In VRDs, visual and layout information is critical for document understanding, and texts in such documents cannot be serialized into the one-dimensional sequence without losing information. Classic information extraction models such as BiLSTM-CRF typically operate on text sequences and do not incorporate visual features. In this paper, we introduce a graph convolution based model to combine textual and visual information presented in VRDs. Graph embeddings are trained to summarize the context of a text segment in the document, and further combined with text embeddings for entity extraction. Extensive experiments have been conducted to show that our method outperforms BiLSTM-CRF baselines by significant margins, on two real-world datasets. Additionally, ablation studies are also performed to evaluate the effectiveness of each component of our model.