Recently, O1-like models have emerged as representative examples, illustrating the effectiveness of long chain-of-thought (CoT) in reasoning tasks such as math and coding tasks. In this paper, we introduce DRT-o1, an attempt to bring the success of long CoT to neural machine translation (MT). Specifically, in view of the literature books that might involve similes and metaphors, translating these texts to a target language is very difficult in practice due to cultural differences. In such cases, literal translation often fails to convey the intended meaning effectively. Even for professional human translators, considerable thought must be given to preserving semantics throughout the translation process. To simulate LLMs' long thought ability in MT, we first mine sentences containing similes or metaphors from existing literature books, and then develop a multi-agent framework to translate these sentences via long thought. In the multi-agent framework, a translator is used to iteratively translate the source sentence under the suggestions provided by an advisor. To ensure the effectiveness of the long thoughts, an evaluator is also employed to quantify the translation in each round. In this way, we collect tens of thousands of long-thought MT data, which is used to train our DRT-o1. Using Qwen2.5 and LLama-3.1 as the backbones, DRT-o1 models can learn the thought process during machine translation, and outperform vanilla LLMs as well as existing O1-like LLMs, showing their effectiveness The project is available at https://github.com/krystalan/DRT-o1
Multilingual language models (MLLMs) are crucial for handling text across various languages, yet they often show performance disparities due to differences in resource availability and linguistic characteristics. While the impact of pre-train data percentage and model size on performance is well-known, our study reveals additional critical factors that significantly influence MLLM effectiveness. Analyzing a wide range of features, including geographical, linguistic, and resource-related aspects, we focus on the SIB-200 dataset for classification and the Flores-200 dataset for machine translation, using regression models and SHAP values across 204 languages. Our findings identify token similarity and country similarity as pivotal factors, alongside pre-train data and model size, in enhancing model performance. Token similarity facilitates cross-lingual transfer, while country similarity highlights the importance of shared cultural and linguistic contexts. These insights offer valuable guidance for developing more equitable and effective multilingual language models, particularly for underrepresented languages.
As the Information Retrieval (IR) field increasingly recognizes the importance of inclusivity, addressing the needs of low-resource languages remains a significant challenge. This paper introduces the first large-scale Urdu IR dataset, created by translating the MS MARCO dataset through machine translation. We establish baseline results through zero-shot learning for IR in Urdu and subsequently apply the mMARCO multilingual IR methodology to this newly translated dataset. Our findings demonstrate that the fine-tuned model (Urdu-mT5-mMARCO) achieves a Mean Reciprocal Rank (MRR@10) of 0.247 and a Recall@10 of 0.439, representing significant improvements over zero-shot results and showing the potential for expanding IR access for Urdu speakers. By bridging access gaps for speakers of low-resource languages, this work not only advances multilingual IR research but also emphasizes the ethical and societal importance of inclusive IR technologies. This work provides valuable insights into the challenges and solutions for improving language representation and lays the groundwork for future research, especially in South Asian languages, which can benefit from the adaptable methods used in this study.
We introduce MT-LENS, a framework designed to evaluate Machine Translation (MT) systems across a variety of tasks, including translation quality, gender bias detection, added toxicity, and robustness to misspellings. While several toolkits have become very popular for benchmarking the capabilities of Large Language Models (LLMs), existing evaluation tools often lack the ability to thoroughly assess the diverse aspects of MT performance. MT-LENS addresses these limitations by extending the capabilities of LM-eval-harness for MT, supporting state-of-the-art datasets and a wide range of evaluation metrics. It also offers a user-friendly platform to compare systems and analyze translations with interactive visualizations. MT-LENS aims to broaden access to evaluation strategies that go beyond traditional translation quality evaluation, enabling researchers and engineers to better understand the performance of a NMT model and also easily measure system's biases.
In this paper, we investigate the role of attention heads in Context-aware Machine Translation models for pronoun disambiguation in the English-to-German and English-to-French language directions. We analyze their influence by both observing and modifying the attention scores corresponding to the plausible relations that could impact a pronoun prediction. Our findings reveal that while some heads do attend the relations of interest, not all of them influence the models' ability to disambiguate pronouns. We show that certain heads are underutilized by the models, suggesting that model performance could be improved if only the heads would attend one of the relations more strongly. Furthermore, we fine-tune the most promising heads and observe the increase in pronoun disambiguation accuracy of up to 5 percentage points which demonstrates that the improvements in performance can be solidified into the models' parameters.
Large language models (LLMs) have shown superior capabilities in translating figurative language compared to neural machine translation (NMT) systems. However, the impact of different prompting methods and LLM-NMT combinations on idiom translation has yet to be thoroughly investigated. This paper introduces two parallel datasets of sentences containing idiomatic expressions for Persian$\rightarrow$English and English$\rightarrow$Persian translations, with Persian idioms sampled from our PersianIdioms resource, a collection of 2,200 idioms and their meanings. Using these datasets, we evaluate various open- and closed-source LLMs, NMT models, and their combinations. Translation quality is assessed through idiom translation accuracy and fluency. We also find that automatic evaluation methods like LLM-as-a-judge, BLEU and BERTScore are effective for comparing different aspects of model performance. Our experiments reveal that Claude-3.5-Sonnet delivers outstanding results in both translation directions. For English$\rightarrow$Persian, combining weaker LLMs with Google Translate improves results, while Persian$\rightarrow$English translations benefit from single prompts for simpler models and complex prompts for advanced ones.
This research builds upon the Latvian Twitter Eater Corpus (LTEC), which is focused on the narrow domain of tweets related to food, drinks, eating and drinking. LTEC has been collected for more than 12 years and reaching almost 3 million tweets with the basic information as well as extended automatically and manually annotated metadata. In this paper we supplement the LTEC with manually annotated subsets of evaluation data for machine translation, named entity recognition, timeline-balanced sentiment analysis, and text-image relation classification. We experiment with each of the data sets using baseline models and highlight future challenges for various modelling approaches.
Retrieval-augmented generation (RAG) introduces additional information to enhance large language models (LLMs). In machine translation (MT), previous work typically retrieves in-context examples from paired MT corpora, or domain-specific knowledge from knowledge graphs, to enhance models' MT ability. However, a large amount of world knowledge is organized in unstructured documents, and might not be fully paired across different languages. In this paper, we study retrieval-augmented MT using unstructured documents. Specifically, we build RAGtrans, the first benchmark to train and evaluate LLMs' retrieval-augmented MT ability. RAGtrans contains 79K MT samples collected via GPT-4o and human translators. Besides, documents from different languages are also provided to supply the knowledge to these samples. Based on RAGtrans, we further propose a multi-task training method to teach LLMs how to use information from multilingual documents during their translation. The method uses existing multilingual corpora to create auxiliary training objectives without additional labeling requirements. Extensive experiments show that the method improves LLMs by 1.58-3.09 BLEU and 1.00-2.03 COMET scores.
This paper presents a multi-way parallel English-Tamil-Sinhala corpus annotated with Named Entities (NEs), where Sinhala and Tamil are low-resource languages. Using pre-trained multilingual Language Models (mLMs), we establish new benchmark Named Entity Recognition (NER) results on this dataset for Sinhala and Tamil. We also carry out a detailed investigation on the NER capabilities of different types of mLMs. Finally, we demonstrate the utility of our NER system on a low-resource Neural Machine Translation (NMT) task. Our dataset is publicly released: https://github.com/suralk/multiNER.
Existing multilingual neural machine translation (MNMT) approaches mainly focus on improving models with the encoder-decoder architecture to translate multiple languages. However, decoder-only architecture has been explored less in MNMT due to its underperformance when trained on parallel data solely. In this work, we attribute the issue of the decoder-only architecture to its lack of language transfer capability. Specifically, the decoder-only architecture is insufficient in encoding source tokens with the target language features. We propose dividing the decoding process into two stages so that target tokens are explicitly excluded in the first stage to implicitly boost the transfer capability across languages. Additionally, we impose contrastive learning on translation instructions, resulting in improved performance in zero-shot translation. We conduct experiments on TED-19 and OPUS-100 datasets, considering both training from scratch and fine-tuning scenarios. Experimental results show that, compared to the encoder-decoder architecture, our methods not only perform competitively in supervised translations but also achieve improvements of up to 3.39 BLEU, 6.99 chrF++, 3.22 BERTScore, and 4.81 COMET in zero-shot translations.
Many of the world's languages have insufficient data to train high-performing general neural machine translation (NMT) models, let alone domain-specific models, and often the only available parallel data are small amounts of religious texts. Hence, domain adaptation (DA) is a crucial issue faced by contemporary NMT and has, so far, been underexplored for low-resource languages. In this paper, we evaluate a set of methods from both low-resource NMT and DA in a realistic setting, in which we aim to translate between a high-resource and a low-resource language with access to only: a) parallel Bible data, b) a bilingual dictionary, and c) a monolingual target-domain corpus in the high-resource language. Our results show that the effectiveness of the tested methods varies, with the simplest one, DALI, being most effective. We follow up with a small human evaluation of DALI, which shows that there is still a need for more careful investigation of how to accomplish DA for low-resource NMT.
Transformer neural networks (TNN) excel in natural language processing (NLP), machine translation, and computer vision (CV) without relying on recurrent or convolutional layers. However, they have high computational and memory demands, particularly on resource-constrained devices like FPGAs. Moreover, transformer models vary in processing time across applications, requiring custom models with specific parameters. Designing custom accelerators for each model is complex and time-intensive. Some custom accelerators exist with no runtime adaptability, and they often rely on sparse matrices to reduce latency. However, hardware designs become more challenging due to the need for application-specific sparsity patterns. This paper introduces ADAPTOR, a runtime-adaptive accelerator for dense matrix computations in transformer encoders and decoders on FPGAs. ADAPTOR enhances the utilization of processing elements and on-chip memory, enhancing parallelism and reducing latency. It incorporates efficient matrix tiling to distribute resources across FPGA platforms and is fully quantized for computational efficiency and portability. Evaluations on Xilinx Alveo U55C data center cards and embedded platforms like VC707 and ZCU102 show that our design is 1.2$\times$ and 2.87$\times$ more power efficient than the NVIDIA K80 GPU and the i7-8700K CPU respectively. Additionally, it achieves a speedup of 1.7 to 2.25$\times$ compared to some state-of-the-art FPGA-based accelerators.
This study presents the development of a part-of-speech (POS) tagging model to extract the skeletal structure of sentences using transfer learning with the BERT architecture for token classification. The model, fine-tuned on Russian text, demonstrating its effectiveness. The approach offers potential applications in enhancing natural language processing tasks, such as improving machine translation. Keywords: part of speech tagging, morphological analysis, natural language processing, BERT.
This study presents a comprehensive evaluation of GPT-4's translation capabilities compared to human translators of varying expertise levels. Through systematic human evaluation using the MQM schema, we assess translations across three language pairs (Chinese$\longleftrightarrow$English, Russian$\longleftrightarrow$English, and Chinese$\longleftrightarrow$Hindi) and three domains (News, Technology, and Biomedical). Our findings reveal that GPT-4 achieves performance comparable to junior-level translators in terms of total errors, while still lagging behind senior translators. Unlike traditional Neural Machine Translation systems, which show significant performance degradation in resource-poor language directions, GPT-4 maintains consistent translation quality across all evaluated language pairs. Through qualitative analysis, we identify distinctive patterns in translation approaches: GPT-4 tends toward overly literal translations and exhibits lexical inconsistency, while human translators sometimes over-interpret context and introduce hallucinations. This study represents the first systematic comparison between LLM and human translators across different proficiency levels, providing valuable insights into the current capabilities and limitations of LLM-based translation systems.
Neural Machine Translation systems are used in diverse applications due to their impressive performance. However, recent studies have shown that these systems are vulnerable to carefully crafted small perturbations to their inputs, known as adversarial attacks. In this paper, we propose a new type of adversarial attack against NMT models. In this attack, we find a word to be added between two sentences such that the second sentence is ignored and not translated by the NMT model. The word added between the two sentences is such that the whole adversarial text is natural in the source language. This type of attack can be harmful in practical scenarios since the attacker can hide malicious information in the automatic translation made by the target NMT model. Our experiments show that different NMT models and translation tasks are vulnerable to this type of attack. Our attack can successfully force the NMT models to ignore the second part of the input in the translation for more than 50% of all cases while being able to maintain low perplexity for the whole input.
This study focuses on recognizing Bangladeshi dialects and converting diverse Bengali accents into standardized formal Bengali speech. Dialects, often referred to as regional languages, are distinctive variations of a language spoken in a particular location and are identified by their phonetics, pronunciations, and lexicon. Subtle changes in pronunciation and intonation are also influenced by geographic location, educational attainment, and socioeconomic status. Dialect standardization is needed to ensure effective communication, educational consistency, access to technology, economic opportunities, and the preservation of linguistic resources while respecting cultural diversity. Being the fifth most spoken language with around 55 distinct dialects spoken by 160 million people, addressing Bangla dialects is crucial for developing inclusive communication tools. However, limited research exists due to a lack of comprehensive datasets and the challenges of handling diverse dialects. With the advancement in multilingual Large Language Models (mLLMs), emerging possibilities have been created to address the challenges of dialectal Automated Speech Recognition (ASR) and Machine Translation (MT). This study presents an end-to-end pipeline for converting dialectal Noakhali speech to standard Bangla speech. This investigation includes constructing a large-scale diverse dataset with dialectal speech signals that tailored the fine-tuning process in ASR and LLM for transcribing the dialect speech to dialect text and translating the dialect text to standard Bangla text. Our experiments demonstrated that fine-tuning the Whisper ASR model achieved a CER of 0.8% and WER of 1.5%, while the BanglaT5 model attained a BLEU score of 41.6% for dialect-to-standard text translation.
Reranking a list of candidates from a machine translation system with an external scoring model and returning the highest-scoring candidate remains a simple and effective method for improving the overall output quality. Translation scoring models continue to grow in size, with the best models being comparable to generation models. Thus, reranking can add substantial computational cost to the translation pipeline. In this work, we pose reranking as a Bayesian optimization (BayesOpt) problem. By strategically selecting candidates to score based on a balance of exploration and exploitation, we show that it is possible to find top-scoring candidates when scoring only a fraction of the candidate list. For instance, our method achieves the same CometKiwi score using only 70 scoring evaluations compared a baseline system using 180. We present a multi-fidelity setting for BayesOpt, where the candidates are first scored with a cheaper but noisier proxy scoring model, which further improves the cost-performance tradeoff when using smaller but well-trained distilled proxy scorers.
Deep learning underpins most of the currently advanced natural language processing (NLP) tasks such as textual classification, neural machine translation (NMT), abstractive summarization and question-answering (QA). However, the robustness of the models, particularly QA models, against adversarial attacks is a critical concern that remains insufficiently explored. This paper introduces QA-Attack (Question Answering Attack), a novel word-level adversarial strategy that fools QA models. Our attention-based attack exploits the customized attention mechanism and deletion ranking strategy to identify and target specific words within contextual passages. It creates deceptive inputs by carefully choosing and substituting synonyms, preserving grammatical integrity while misleading the model to produce incorrect responses. Our approach demonstrates versatility across various question types, particularly when dealing with extensive long textual inputs. Extensive experiments on multiple benchmark datasets demonstrate that QA-Attack successfully deceives baseline QA models and surpasses existing adversarial techniques regarding success rate, semantics changes, BLEU score, fluency and grammar error rate.
We introduce the GEM (Generative Estimator for Mutual Information), an evaluation metric for assessing language generation by Large Language Models (LLMs), particularly in generating informative judgments, without the need for a gold standard reference. GEM broadens the scenarios where we can benchmark LLM generation performance-from traditional ones, like machine translation and summarization, where gold standard references are readily available, to subjective tasks without clear gold standards, such as academic peer review. GEM uses a generative model to estimate mutual information between candidate and reference responses, without requiring the reference to be a gold standard. In experiments on a human-annotated dataset, GEM demonstrates competitive correlations with human scores compared to the state-of-the-art GPT-4o Examiner, and outperforms all other baselines. Additionally, GEM is more robust against strategic manipulations, such as rephrasing or elongation, which can artificially inflate scores under a GPT-4o Examiner. We also present GRE-bench (Generating Review Evaluation Benchmark) which evaluates LLMs based on how well they can generate high-quality peer reviews for academic research papers. Because GRE-bench is based upon GEM, it inherits its robustness properties. Additionally, GRE-bench circumvents data contamination problems (or data leakage) by using the continuous influx of new open-access research papers and peer reviews each year. We show GRE-bench results of various popular LLMs on their peer review capabilities using the ICLR2023 dataset.
In translation, a concept represented by a single word in a source language can have multiple variations in a target language. The task of lexical selection requires using context to identify which variation is most appropriate for a source text. We work with native speakers of nine languages to create DTAiLS, a dataset of 1,377 sentence pairs that exhibit cross-lingual concept variation when translating from English. We evaluate recent LLMs and neural machine translation systems on DTAiLS, with the best-performing model, GPT-4, achieving from 67 to 85% accuracy across languages. Finally, we use language models to generate English rules describing target-language concept variations. Providing weaker models with high-quality lexical rules improves accuracy substantially, in some cases reaching or outperforming GPT-4.