The paper explores how to improve literary machine translation by using datasets that contain multiple valid translations of the same source text. It introduces a filtering framework that selects source texts whose references show meaningful variation while staying faithful, based on semantic similarity. Experiments show that fine‑tuning on medium to high similarity data outperforms low similarity data, and that using only this filtered subset can match or exceed performance achieved with the full unfiltered set. Additionally, the study compares synthetic translations generated by large language models with human expert translations, finding that fine‑tuning on human expert data yields better results in both automatic metrics and human evaluations, underscoring the continued importance of expert translations for literary MT.
arXiv:2606. 05924v1 Announce Type: cross Abstract: Literary translation poses unique challenges due to the scarcity of high-quality annotated data and the need to balance expression fluency with literary effect.
arXiv:2509. 07829v4 Announce Type: replace-cross Abstract: Literary translation has recently gained attention as a distinct and complex task in machine translation research, yet translation by small open models remains an open problem, particularly for low-resource languages such as Romanian.
arXiv:2606.26040v2 Announce Type: replace Abstract: AI translation of literary works is increasingly common. While the content may be rendered adequately, we do not know enough about how readers expe...
arXiv:2606. 05444v1 Announce Type: cross Abstract: Coreference resolution is a core NLP task, having a broad range of downstream applications, e.
The paper introduces Loci Similes, a benchmark for detecting intertextual links in Latin literature. It provides a curated dataset of about 176,000 text segments and 1,490 expert-verified parallels, including 945 labeled references from an existing source. Baselines for retrieval and classification are established using both lexical methods and pretrained encoder language models.
The paper introduces the Last Translation Benchmark (LTB), a live dataset of human-authored and peer‑reviewed examples—including texts, images, audio, and videos—that are designed to break current state‑of‑the‑art machine translation models. Each example is accompanied by handcrafted verification rules that specify concrete failure cases, providing a reliable and actionable evaluation method. The benchmark aims to overcome the limitations of existing automatic metrics and gold human evaluations, which often lack reproducibility, objectivity, and scalability.
arXiv:2609.22805v1 Announce Type: new Abstract: Natural language generation (NLG) tasks span the spectrum of conditional entropy, ranging from highly constrained machine translation to open-ended dia...
The paper investigates how machine‑translated English data from 24 diverse source languages influences small English language models. It finds that source language affects model behavior: lexical diversity drives overall perplexity, while grammatical performance correlates with typological similarity to English when sufficient data is used. Additionally, translation quality strongly predicts language‑modeling performance.
The paper investigates how machine translation can be tailored to specific audiences and intents, a capability enabled by large language models (LLMs). By systematically evaluating purpose-driven MT across 50 languages, 5 model sizes, and 8 text domains, the authors find that explicit instructions significantly improve translation adaptiveness, especially for informal domains, larger models, and higher-resource languages. They also show that traditional MT metrics often penalize adapted translations and that models can self-generate useful instructions from context, closing a large portion of the adaptiveness gap.
The paper examines the intermediate‑entropy regime of natural language generation, where a single input can have multiple valid outputs such as in peer‑review generation and image captioning. It introduces a target‑side consolidation method that merges all references into a single target preserving every distinct information unit, and presents the MERC‑67K dataset of 36,537 papers and 31,014 images with both original and consolidated references. Human evaluations show that models trained on consolidated references achieve higher coverage and overall quality scores, outperforming multi‑reference baselines across several large language models while using fewer individual references during training.
arXiv:2512. 07540v4 Announce Type: replace-cross Abstract: Error Span Detection (ESD) extends automatic machine translation (MT) evaluation by localizing translation errors and labeling their severity.
The paper introduces a semiotics-aware framework for assessing natural language generation, focusing on how well two texts align in terms of contextual meaning and discourse references. It defines two metrics—Semiotic Fidelity and Semiotic Coverage—to quantify how much of one text’s semiotic profile is supported by the other and how much of the other’s profile is recovered. Experiments reveal that coverage is usually lower than fidelity, and that language models align best with human-curated data at low sampling temperatures, with higher temperatures diminishing this alignment.