ABAlexandra Birch

Alexandra Birch

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Alexandra Birch has a total of 75 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Top Cited Papers
    Open Access

    Machine Translation in the Era of Large Language Models:A Survey of Historical and Emerging Problems

    by Duygu Ataman, Alexandra Birch, Nizar Habash, Marcello Federico, Philipp Koehn, Kyunghyun Cho

    Historically regarded as one of the most challenging tasks presented to achieve complete artificial intelligence (AI), machine translation (MT) research has seen continuous devotion over the past decade, resulting in cutting-edge architectures for the modeling of sequential information. While the majority of statistical models traditionally relied on the idea of learning from parallel translation examples, recent research exploring self-supervised and multi-task learning methods extended the capabilities of MT models, eventually allowing the creation of general-purpose large language models (LLMs). In addition to versatility in providing translations useful across languages and domains, LLMs can in principle perform any natural language processing (NLP) task given sufficient amount of task-specific examples. While LLMs now reach a point where they can both replace and augment traditional MT models, the extent of their advantages and the ways in which they leverage translation capabilities across multilingual NLP tasks remains a wide area for exploration. In this literature survey, we present an introduction to the current position of MT research with a historical look at different modeling approaches to MT, how these might be advantageous for the solution of particular problems, and which problems are solved or remain open in regard to recent developments. We also discuss the connection of MT models leading to the development of prominent LLM architectures, how they continue to support LLM performance across different tasks by providing a means for cross-lingual knowledge transfer, and the redefinition of the task with the possibilities that LLM technology brings.


  • Open Access

    The Prosody of Emojis

    by Giulio Zhou, Tsz Kin Lam, Alexandra Birch, Barry Haddow

    Prosodic features such as pitch, timing, and intonation are central to spoken communication, conveying emotion, intent, and discourse structure. In text-based settings, where these cues are absent, emojis act as visual surrogates that add affective and pragmatic nuance. This study examines how emojis influence prosodic realisation in speech and how listeners interpret prosodic cues to recover emoji meanings. Unlike previous work, we directly link prosody and emoji by analysing actual human speech data, collected through structured but open-ended production and perception tasks. This provides empirical evidence of how emoji semantics shape spoken delivery and perception. Results show that speakers adapt their prosody based on emoji cues, listeners can often identify the intended emoji from prosodic variation alone, and greater semantic differences between emojis correspond to increased prosodic divergence. These findings suggest that emojis can act as meaningful carriers of prosodic intent, offering insight into their communicative role in digitally mediated contexts.


  • Open Access

    Generalizing From Short to Long: Effective Data Synthesis for Long-Context Instruction Tuning

    by Wenhao Zhu, Pinzhen Chen, Hanxu Hu, Shujian Huang, Fei Yuan, Jiajun Chen, Alexandra Birch

    Long-context modelling for large language models (LLMs) has been a key area of recent research because many real world use cases require reasoning over longer inputs such as documents. The focus of research into modelling long context has been on how to model position and there has been little investigation into other important aspects of language modelling such as instruction tuning. Long context training examples are challenging and expensive to create and use. In this paper, we investigate how to design instruction data for the post-training phase of a long context pre-trained model: how much and what type of context is needed for optimal and efficient post-training. Our controlled study reveals that models instruction-tuned on short contexts can effectively generalize to longer ones, while also identifying other critical factors such as instruction difficulty and context composition. Based on these findings, we propose context synthesis, a novel data synthesis framework that leverages off-the-shelf LLMs to generate extended background contexts for high-quality instruction-answer pairs. Experiment results on the document-level benchmark (LongBench) demonstrate that our proposed approach outperforms previous instruction synthesis approaches and comes close to the performance of human-annotated long-context instruction data. The project will be available at: https://github.com/NJUNLP/context-synthesis.

  • Open Access

    Demystifying Multilingual Reasoning in Process Reward Modeling

    by Weixuan Wang, Minghao Wu, Barry Haddow, Alexandra Birch

    Weixuan Wang, Minghao Wu, Barry Haddow, Alexandra Birch. Findings of the Association for Computational Linguistics: EMNLP 2025. 2025.


  • Open Access

    Bridging the Language Gaps in Large Language Models with Inference-Time Cross-Lingual Intervention

    by Weixuan Wang, Minghao Wu, Barry Haddow, Alexandra Birch

    Weixuan Wang, Minghao Wu, Barry Haddow, Alexandra Birch. Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2025.


  • Open Access

    Generics are puzzling. Can language models find the missing piece?

    by Gustavo Cilleruelo Calderón, Emily Allaway, Barry Haddow, Alexandra Birch

    Generic sentences express generalisations about the world without explicit quantification. Although generics are central to everyday communication, building a precise semantic framework has proven difficult, in part because speakers use generics to generalise properties with widely different statistical prevalence. In this work, we study the implicit quantification and context-sensitivity of generics by leveraging language models as models of language. We create ConGen, a dataset of 2873 naturally occurring generic and quantified sentences in context, and define p-acceptability, a metric based on surprisal that is sensitive to quantification. Our experiments show generics are more context-sensitive than determiner quantifiers and about 20% of naturally occurring generics we analyze express weak generalisations. We also explore how human biases in stereotypes can be observed in language models.


  • Open Access

    Bridging the Language Gaps in Large Language Models with Inference-Time Cross-Lingual Intervention

    by Weixuan Wang, Minghao Wu, Barry Haddow, Alexandra Birch

    Large Language Models (LLMs) have shown remarkable capabilities in natural language processing but exhibit significant performance gaps among different languages. Most existing approaches to address these disparities rely on pretraining or fine-tuning, which are resource-intensive. To overcome these limitations without incurring significant costs, we propose Inference-Time Cross-Lingual Intervention (INCLINE), a novel framework that enhances LLM performance on low-performing (source) languages by aligning their internal representations with those of high-performing (target) languages during inference. INCLINE initially learns alignment matrices using parallel sentences from source and target languages through a Least-Squares optimization, and then applies these matrices during inference to transform the low-performing language representations toward the high-performing language space. Extensive experiments on nine benchmarks with five LLMs demonstrate that INCLINE significantly improves performance across diverse tasks and languages, compared to recent strong baselines. Our analysis demonstrates that INCLINE is highly cost-effective and applicable to a wide range of applications. In addition, we release the code to foster research along this line: https://github.com/weixuan-wang123/INCLINE.

  • Open Access

    Can GPT-3.5 generate and code discharge summaries?

    by Matúš Falis, Aryo Pradipta Gema, Hang Dong, Luke Daines, Siddharth Basetti, Michael Holder, Rose S Penfold, Alexandra Birch, Beatrice Alex

    Objectives: The aim of this study was to investigate GPT-3.5 in generating and coding medical documents with International Classification of Diseases (ICD)-10 codes for data augmentation on low-resource labels. Materials and Methods: Employing GPT-3.5 we generated and coded 9606 discharge summaries based on lists of ICD-10 code descriptions of patients with infrequent (or generation) codes within the MIMIC-IV dataset. Combined with the baseline training set, this formed an augmented training set. Neural coding models were trained on baseline and augmented data and evaluated on an MIMIC-IV test set. We report micro- and macro-F1 scores on the full codeset, generation codes, and their families. Weak Hierarchical Confusion Matrices determined within-family and outside-of-family coding errors in the latter codesets. The coding performance of GPT-3.5 was evaluated on prompt-guided self-generated data and real MIMIC-IV data. Clinicians evaluated the clinical acceptability of the generated documents. Results: Data augmentation results in slightly lower overall model performance but improves performance for the generation candidate codes and their families, including 1 absent from the baseline training data. Augmented models display lower out-of-family error rates. GPT-3.5 identifies ICD-10 codes by their prompted descriptions but underperforms on real data. Evaluators highlight the correctness of generated concepts while suffering in variety, supporting information, and narrative. Discussion and Conclusion: While GPT-3.5 alone given our prompt setting is unsuitable for ICD-10 coding, it supports data augmentation for training neural models. Augmentation positively affects generation code families but mainly benefits codes with existing examples. Augmentation reduces out-of-family errors. Documents generated by GPT-3.5 state prompted concepts correctly but lack variety, and authenticity in narratives.


  • Open Access

    Cultural Adaptation of Menus: A Fine-Grained Approach

    by Zhonghe Zhang, Xiaoyu He, Vivek Iyer, Alexandra Birch

    Machine Translation of Culture-Specific Items (CSIs) poses significant challenges. Recent work on CSI translation has shown some success using Large Language Models (LLMs) to adapt to different languages and cultures; however, a deeper analysis is needed to examine the benefits and pitfalls of each method. In this paper, we introduce the ChineseMenuCSI dataset, the largest for Chinese-English menu corpora, annotated with CSI vs Non-CSI labels and a fine-grained test set. We define three levels of CSI figurativeness for a more nuanced analysis and develop a novel methodology for automatic CSI identification, which outperforms GPT-based prompts in most categories. Importantly, we are the first to integrate human translation theories into LLM-driven translation processes, significantly improving translation accuracy, with COMET scores increasing by up to 7 points.

  • Open Access

    Optimising Calls to Large Language Models with Uncertainty-Based Two-Tier Selection

    by Guillem Ramírez, Alexandra Birch, Ivan Titov

    Researchers and practitioners operating on a limited budget face the cost-performance trade-off dilemma. The challenging decision often centers on whether to use a large LLM with better performance or a smaller one with reduced costs. This has motivated recent research in the optimisation of LLM calls. Either a cascading strategy is used, where a smaller LLM or both are called sequentially, or a routing strategy is used, where only one model is ever called. Both scenarios are dependent on a decision criterion which is typically implemented by an extra neural model. In this work, we propose a simpler solution; we use only the uncertainty of the generations of the small LLM as the decision criterion. We compare our approach with both cascading and routing strategies using three different pairs of pre-trained small and large LLMs, on nine different tasks and against approaches that require an additional neural model. Our experiments reveal this simple solution optimally balances cost and performance, outperforming existing methods on 25 out of 27 experimental setups.