Showing 1–26 of 26 results for author: Buttery, P

LangMAP: A Language-Adaptive Approach to Tokenization

Authors: Clara Meister, Suchir Salhan, Andrzej Szablewski, Pietro Lesci, Paula Buttery, Tiago Pimentel

Abstract: Language-specific tokenizers improve tokenization quality and the downstream performance of models on those languages. However, using such a tokenizer comes at a cost: either a new model must be trained from scratch, or the vocabulary of an existing pretrained model must be adapted. We propose Language-adaptive Maximum a Posteriori (LangMAP) Tokenization, a tokenization scheme that extends the Uni… ▽ More Language-specific tokenizers improve tokenization quality and the downstream performance of models on those languages. However, using such a tokenizer comes at a cost: either a new model must be trained from scratch, or the vocabulary of an existing pretrained model must be adapted. We propose Language-adaptive Maximum a Posteriori (LangMAP) Tokenization, a tokenization scheme that extends the UnigramLM algorithm to the multilingual setting, producing language-specific tokenization from a single shared vocabulary. Notably, LangMAP can be used when training a multilingual language model from scratch or to adapt a pretrained model's tokenizer to individual languages without changing its vocabulary. While language labels are required at training time, a key feature of the algorithm is that it then performs language-specific tokenization at inference without knowledge of the input's language. Across 14 open-source tokenizers, 9 natural languages, and 9 programming languages, LangMAP improves morphological boundary alignment and, for all coding languages tested, alignment with abstract syntax tree (AST) leaf boundaries. In fine-tuning experiments, results are mixed: LangMAP improves target-language grammatical acceptability (MultiBLiMP) on the languages tested; its benefits are less consistent on knowledge-related tasks (Global-PIQA, Belebele). △ Less

Submitted 23 June, 2026; v1 submitted 22 June, 2026; originally announced June 2026.

LLMs Contain Multitudes: How Deployment Context Reshapes Model-Level Preferences and Values

Authors: Filip Trhlik, Aoife O'Flynn, Angela Yu, Arduin Findeis, Paula Buttery

Abstract: Large language models (LLMs) are increasingly characterised in recent evaluation work as having stable, model-level preference and value systems. However, accompanying robustness checks are limited to incidental prompt perturbations such as syntax variation and option reordering. This leaves open whether the measured properties survive when the surrounding task context changes, as it does in most… ▽ More Large language models (LLMs) are increasingly characterised in recent evaluation work as having stable, model-level preference and value systems. However, accompanying robustness checks are limited to incidental prompt perturbations such as syntax variation and option reordering. This leaves open whether the measured properties survive when the surrounding task context changes, as it does in most real deployments. We test this directly across two established pairwise paradigms: ranking country preferences and eliciting utility judgements. In both, we make the deployment context -- the high-level task the model is performing while making concrete value-dependent choices -- our controlled variable, varied across framings such as writing a Reddit post or a news article. Across five LLMs and over 1.2M pairwise decisions, deployment context produces variation far larger than prompt paraphrasing and temperature controls. In country preference rankings over 15 countries, context induces widespread, statistically significant rank shifts; the aggregate Global North favouritism reported in prior work is itself context-dependent, with each model's bias shifting systematically across contexts. In utility elicitation over 50 outcomes, broad cross-category ordering is preserved, but fine-grained rankings within domains vary substantially, and cardinal exchange rates between outcomes (e.g. how many lives in one region equal one in another) shift by a factor of 2.47 at the median. Reported model-level preferences and utilities are therefore better understood as context-conditioned measurements than fixed model-level properties: safety guarantees obtained under one framing provide limited assurance in another. △ Less

Submitted 11 June, 2026; originally announced June 2026.

Comments: 68 pages, 54 figures, 54 tables

Bias Dynamics in BabyLMs: Towards a Compute-Efficient Sandbox for Democratising Pre-Training Debiasing

Authors: Filip Trhlik, Andrew Caines, Paula Buttery

Abstract: Pre-trained language models (LMs) have, over the last few years, grown substantially in both societal adoption and training costs. This rapid growth in size has constrained progress in understanding and mitigating their biases. Since re-training LMs is prohibitively expensive, most debiasing work has focused on post-hoc or masking-based strategies, which often fail to address the underlying causes… ▽ More Pre-trained language models (LMs) have, over the last few years, grown substantially in both societal adoption and training costs. This rapid growth in size has constrained progress in understanding and mitigating their biases. Since re-training LMs is prohibitively expensive, most debiasing work has focused on post-hoc or masking-based strategies, which often fail to address the underlying causes of bias. In this work, we seek to democratise pre-model debiasing research by using low-cost proxy models. Specifically, we investigate BabyLMs, compact BERT-like models trained on small and mutable corpora that can approximate bias acquisition and learning dynamics of larger models. We show that BabyLMs display closely aligned patterns of intrinsic bias formation and performance development compared to standard BERT models, despite their drastically reduced size. Furthermore, correlations between BabyLMs and BERT hold across multiple intra-model and post-model debiasing methods. Leveraging these similarities, we conduct pre-model debiasing experiments with BabyLMs, replicating prior findings and presenting new insights regarding the influence of gender imbalance and toxicity on bias formation. Our results demonstrate that BabyLMs can serve as an effective sandbox for large-scale LMs, reducing pre-training costs from over 500 GPU-hours to under 30 GPU-hours. This provides a way to democratise pre-model debiasing research and enables faster, more accessible exploration of methods for building fairer LMs. △ Less

Submitted 15 January, 2026; v1 submitted 14 January, 2026; originally announced January 2026.

Comments: 21 pages, 18 figures

Teacher Demonstrations in a BabyLM's Zone of Proximal Development for Contingent Multi-Turn Interaction

Authors: Suchir Salhan, Hongyi Gu, Donya Rooein, Diana Galvan-Sosa, Gabrielle Gaudeau, Andrew Caines, Zheng Yuan, Paula Buttery

Abstract: Multi-turn dialogues between a child and a caregiver are characterized by a property called contingency - that is, prompt, direct, and meaningful exchanges between interlocutors. We introduce ContingentChat, a teacher-student framework that benchmarks and improves multi-turn contingency in a BabyLM trained on 100M words. Using a novel alignment dataset for post-training, BabyLM generates responses… ▽ More Multi-turn dialogues between a child and a caregiver are characterized by a property called contingency - that is, prompt, direct, and meaningful exchanges between interlocutors. We introduce ContingentChat, a teacher-student framework that benchmarks and improves multi-turn contingency in a BabyLM trained on 100M words. Using a novel alignment dataset for post-training, BabyLM generates responses that are more grammatical and cohesive. Experiments with adaptive teacher decoding strategies show limited additional gains. ContingentChat demonstrates the benefits of targeted post-training for dialogue quality and indicates that contingency remains a challenging goal for BabyLMs. △ Less

Submitted 23 October, 2025; originally announced October 2025.

Comments: Outstanding Paper Award, EMNLP 2025 BabyLM Workshop - Oral presentation, Suzhou, China

What is the Best Sequence Length for BABYLM?

Authors: Suchir Salhan, Richard Diehl Martinez, Zébulon Goriely, Paula Buttery

Abstract: Transformer language models typically operate with a fixed-length context window, which has grown in step with large-scale pretraining datasets. In the BabyLM Challenge, however, many past submissions have defaulted to using much shorter sequence lengths. We examine the impact of sequence length on BabyLM pretraining, to answer the simple question: what sequence length should we be using when trai… ▽ More Transformer language models typically operate with a fixed-length context window, which has grown in step with large-scale pretraining datasets. In the BabyLM Challenge, however, many past submissions have defaulted to using much shorter sequence lengths. We examine the impact of sequence length on BabyLM pretraining, to answer the simple question: what sequence length should we be using when training Baby LMs? Using 100M-word training data and fixed compute budgets, we compare 125M-parameter Mamba and OPT models, finding that although longer is often better, the optimal length depends on both task and architecture. Shorter sequences are sufficient for grammatical generalization tasks whereas longer contexts benefit morphological analogical reasoning tasks. △ Less

Submitted 22 October, 2025; originally announced October 2025.

Comments: Paper Accepted at the 2025 BabyLM Workshop @ EMNLP (Suzhou, China)

BLiSS 1.0: Evaluating Bilingual Learner Competence in Second Language Small Language Models

Authors: Yuan Gao, Suchir Salhan, Andrew Caines, Paula Buttery, Weiwei Sun

Abstract: To bridge the gap between performance-oriented benchmarks and the evaluation of cognitively inspired models, we introduce BLiSS 1.0, a Benchmark of Learner Interlingual Syntactic Structure. Our benchmark operationalizes a new paradigm of selective tolerance, testing whether a model finds a naturalistic learner error more plausible than a matched, artificial error within the same sentence. Construc… ▽ More To bridge the gap between performance-oriented benchmarks and the evaluation of cognitively inspired models, we introduce BLiSS 1.0, a Benchmark of Learner Interlingual Syntactic Structure. Our benchmark operationalizes a new paradigm of selective tolerance, testing whether a model finds a naturalistic learner error more plausible than a matched, artificial error within the same sentence. Constructed from over 2.8 million naturalistic learner sentences, BLiSS provides 136,867 controlled triplets (corrected, learner, artificial) for this purpose. Experiments on a diverse suite of models demonstrate that selective tolerance is a distinct capability from standard grammaticality, with performance clustering strongly by training paradigm. This validates BLiSS as a robust tool for measuring how different training objectives impact a model's alignment with the systematic patterns of human language acquisition. △ Less

Submitted 22 October, 2025; originally announced October 2025.

Comments: Accepted Paper at the BabyLM Workshop 2025 @ EMNLP (Presentation in Suzhou, China)

Looking to Learn: Token-wise Dynamic Gating for Low-Resource Vision-Language Modelling

Authors: Bianca-Mihaela Ganescu, Suchir Salhan, Andrew Caines, Paula Buttery

Abstract: Training vision-language models on cognitively-plausible amounts of data requires rethinking how models integrate multimodal information. Within the constraints of the Vision track for the BabyLM Challenge 2025, we propose a lightweight decoder-based architecture with (1) token-wise dynamic gating for adaptive fusion of linguistic and visual cues, (2) feature modulation and channel attention to ma… ▽ More Training vision-language models on cognitively-plausible amounts of data requires rethinking how models integrate multimodal information. Within the constraints of the Vision track for the BabyLM Challenge 2025, we propose a lightweight decoder-based architecture with (1) token-wise dynamic gating for adaptive fusion of linguistic and visual cues, (2) feature modulation and channel attention to maximise the utility of limited visual information and (3) auxiliary contrastive objectives for visual grounding. Evaluation on five benchmarks (BLiMP, BLiMP Supplement, EWoK, Winoground and VQA) shows competitive or superior performance to multimodal baselines. More notably, our dynamic gate discovers interpretable patterns without explicit supervision, favouring visual cues for content words and linguistic cues for function words. While we identify limitations in the Challenge constraints, such as the information bottleneck created by global image embeddings and training instability from the dataset split, our findings establish dynamic gating as a powerful tool for efficient multimodal learning, offering both interpretability and performance even under severe constraints. △ Less

Submitted 9 October, 2025; originally announced October 2025.

Comments: Accepted to the EMNLP 2025 BabyLM Workshop

Pico: A Modular Framework for Hypothesis-Driven Small Language Model Research

Authors: Richard Diehl Martinez, David Demitri Africa, Yuval Weiss, Suchir Salhan, Ryan Daniels, Paula Buttery

Abstract: Building language models (LMs), especially small and medium ones, remains more art than science. While large LMs often improve by sheer scale, it is still unclear why many design choices work. For small LMs, this uncertainty is more limiting: tight parameter budgets make each decision critical, yet researchers still lack systematic, scientific ways to test and refine new ideas. We introduce Pico… ▽ More Building language models (LMs), especially small and medium ones, remains more art than science. While large LMs often improve by sheer scale, it is still unclear why many design choices work. For small LMs, this uncertainty is more limiting: tight parameter budgets make each decision critical, yet researchers still lack systematic, scientific ways to test and refine new ideas. We introduce Pico, a lightweight, modular framework that enables systematic, hypothesis-driven research for small and medium-scale language model development. Pico consists of two libraries that together provide a practical sandbox where researchers can make targeted changes to a model's architecture or training procedures and directly observe their effects on the model's behavior. To support reproducible experimentation, we also release a suite of baseline models, pico-decoder, trained under standardized conditions and open-sourced for the community. Case studies highlight how Pico can support iterative small LM design and analysis. △ Less

Submitted 19 September, 2025; originally announced September 2025.

Investigating ReLoRA: Effects on the Learning Dynamics of Small Language Models

Authors: Yuval Weiss, David Demitri Africa, Paula Buttery, Richard Diehl Martinez

Abstract: Parameter-efficient methods like LoRA have revolutionised large language model (LLM) fine-tuning. ReLoRA extends this idea to pretraining by repeatedly merging and reinitialising low-rank adapters, increasing cumulative rank while keeping updates cheap. This aligns well with observations that high-capacity models learn through locally low-rank trajectories that expand over time. By contrast, recen… ▽ More Parameter-efficient methods like LoRA have revolutionised large language model (LLM) fine-tuning. ReLoRA extends this idea to pretraining by repeatedly merging and reinitialising low-rank adapters, increasing cumulative rank while keeping updates cheap. This aligns well with observations that high-capacity models learn through locally low-rank trajectories that expand over time. By contrast, recent work suggests that small language models (SLMs) exhibit rank deficiencies and under-utilise their available dimensionality. This raises a natural question: can ReLoRA's rank-expanding update rule \textit{steer} SLMs toward healthier learning dynamics, mitigating rank bottlenecks in a capacity-constrained regime? We argue SLMs are an ideal testbed: they train quickly, enable controlled ablations, and make rank phenomena more measurable. We present the first systematic study of ReLoRA in SLMs (11M-66M parameters), evaluating both performance and learning dynamics. Across loss, Paloma perplexity, and BLiMP, we find that ReLoRA underperforms full-rank training, with gaps widening at larger scales. Analysis of proportional effective rank and condition numbers shows that ReLoRA amplifies existing rank deficiencies and induces ill-conditioned updates early in training. Our results suggest that while ReLoRA's merge-and-restart strategy can expand ranks in larger models, it does not straightforwardly translate to capacity-limited SLMs, motivating adaptive-rank or hybrid-rank approaches for low-compute pretraining. △ Less

Submitted 2 October, 2025; v1 submitted 16 September, 2025; originally announced September 2025.

Comments: 12 Pages, 6 Tables, 8 Figures

Meta-Pretraining for Zero-Shot Cross-Lingual Named Entity Recognition in Low-Resource Philippine Languages

Authors: David Demitri Africa, Suchir Salhan, Yuval Weiss, Paula Buttery, Richard Diehl Martinez

Abstract: Named-entity recognition (NER) in low-resource languages is usually tackled by finetuning very large multilingual LMs, an option that is often infeasible in memory- or latency-constrained settings. We ask whether small decoder LMs can be pretrained so that they adapt quickly and transfer zero-shot to languages unseen during pretraining. To this end we replace part of the autoregressive objective w… ▽ More Named-entity recognition (NER) in low-resource languages is usually tackled by finetuning very large multilingual LMs, an option that is often infeasible in memory- or latency-constrained settings. We ask whether small decoder LMs can be pretrained so that they adapt quickly and transfer zero-shot to languages unseen during pretraining. To this end we replace part of the autoregressive objective with first-order model-agnostic meta-learning (MAML). Tagalog and Cebuano are typologically similar yet structurally different in their actor/non-actor voice systems, and hence serve as a challenging test-bed. Across four model sizes (11 M - 570 M) MAML lifts zero-shot micro-F1 by 2-6 pp under head-only tuning and 1-3 pp after full tuning, while cutting convergence time by up to 8%. Gains are largest for single-token person entities that co-occur with Tagalog case particles si/ni, highlighting the importance of surface anchors. △ Less

Submitted 4 October, 2025; v1 submitted 2 September, 2025; originally announced September 2025.

Comments: Accepted (poster) to 5th Workshop on Multilingual Representation Learning at EMNLP 2025

Learning Dynamics of Meta-Learning in Small Model Pretraining

Authors: David Demitri Africa, Yuval Weiss, Paula Buttery, Richard Diehl Martinez

Abstract: Large language models are powerful but costly. We ask whether meta-learning can make the pretraining of small language models not only better but also more interpretable. We integrate first-order MAML with subset-masked LM pretraining, producing four LLama-style decoder-only models (11M-570M params), and evaluate it on a fundamental NLP task with many settings and real-world applications. Compared… ▽ More Large language models are powerful but costly. We ask whether meta-learning can make the pretraining of small language models not only better but also more interpretable. We integrate first-order MAML with subset-masked LM pretraining, producing four LLama-style decoder-only models (11M-570M params), and evaluate it on a fundamental NLP task with many settings and real-world applications. Compared with vanilla training, our model (i) reaches the same loss up to 1.6x sooner, (ii) improves F1 on multilingual Universal NER under equal compute, and (iii) makes the training dynamics easy to read: first the network's representations fan out ("diversify") and later they collapse into a smaller, shared subspace ("compress"). This two-stage shift shows up as a rise-and-fall in both effective-rank curves and attention-head entropy. The same curves pinpoint which layers specialise earliest and which later reconverge, giving a compact, interpretable signature of meta-adaptation. Code, checkpoints and WandB logs are released. △ Less

Submitted 6 November, 2025; v1 submitted 4 August, 2025; originally announced August 2025.

Comments: Accepted (oral) to Student Research Workshop at IJCNLP-AACL 2025

ByteSpan: Information-Driven Subword Tokenisation

Authors: Zébulon Goriely, Suchir Salhan, Pietro Lesci, Julius Cheng, Paula Buttery

Abstract: Recent dynamic tokenisation methods operate directly on bytes and pool their latent representations into patches. This bears similarities to computational models of word segmentation that determine lexical boundaries using spikes in an autoregressive model's prediction error. Inspired by this connection, we explore whether grouping predictable bytes - rather than pooling their representations - ca… ▽ More Recent dynamic tokenisation methods operate directly on bytes and pool their latent representations into patches. This bears similarities to computational models of word segmentation that determine lexical boundaries using spikes in an autoregressive model's prediction error. Inspired by this connection, we explore whether grouping predictable bytes - rather than pooling their representations - can yield a useful fixed subword vocabulary. We propose a new information-driven subword tokeniser, ByteSpan, that uses an external byte-level LM during training to identify contiguous predictable byte sequences and group them into subwords. Experiments show that ByteSpan yields efficient vocabularies with higher morphological alignment scores than BPE for English. Multilingual experiments show similar compression and Rényi efficiency for 25 languages. △ Less

Submitted 23 June, 2025; originally announced June 2025.

Comments: Accepted to TokShop 2025 (Non-archival)

BabyLM's First Words: Word Segmentation as a Phonological Probing Task

Authors: Zébulon Goriely, Paula Buttery

Abstract: Language models provide a key framework for studying linguistic theories based on prediction, but phonological analysis using large language models (LLMs) is difficult; there are few phonological benchmarks beyond English and the standard input representation used in LLMs (subwords of graphemes) is not suitable for analyzing the representation of phonemes. In this work, we demonstrate how word seg… ▽ More Language models provide a key framework for studying linguistic theories based on prediction, but phonological analysis using large language models (LLMs) is difficult; there are few phonological benchmarks beyond English and the standard input representation used in LLMs (subwords of graphemes) is not suitable for analyzing the representation of phonemes. In this work, we demonstrate how word segmentation can be used as a phonological probing task, allowing us to study the representations learned by phoneme-based language models trained on child-directed speech across 31 languages. Following computational models of word segmentation, we present unsupervised methods for extracting word boundaries from a trained model using the observation that prediction-error peaks at the start of words. We also use linear probes to identify that these models implicitly track word boundaries, even when they do not appear in training. This cross-lingual work corroborates statistical learning theories of acquisition and empirically motivates new methods for training subword tokenizers. △ Less

Submitted 12 June, 2025; v1 submitted 4 April, 2025; originally announced April 2025.

Comments: Accepted to CoNLL 2025

IPA-CHILDES & G2P+: Feature-Rich Resources for Cross-Lingual Phonology and Phonemic Language Modeling

Authors: Zébulon Goriely, Paula Buttery

Abstract: In this paper, we introduce two resources: (i) G2P+, a tool for converting orthographic datasets to a consistent phonemic representation; and (ii) IPA CHILDES, a phonemic dataset of child-centered speech across 31 languages. Prior tools for grapheme-to-phoneme conversion result in phonemic vocabularies that are inconsistent with established phonemic inventories, an issue which G2P+ addresses by le… ▽ More In this paper, we introduce two resources: (i) G2P+, a tool for converting orthographic datasets to a consistent phonemic representation; and (ii) IPA CHILDES, a phonemic dataset of child-centered speech across 31 languages. Prior tools for grapheme-to-phoneme conversion result in phonemic vocabularies that are inconsistent with established phonemic inventories, an issue which G2P+ addresses by leveraging the inventories in the Phoible database. Using this tool, we augment CHILDES with phonemic transcriptions to produce IPA CHILDES. This new resource fills several gaps in existing phonemic datasets, which often lack multilingual coverage, spontaneous speech, and a focus on child-directed language. We demonstrate the utility of this dataset for phonological research by training phoneme language models on 11 languages and probing them for distinctive features, finding that the distributional properties of phonemes are sufficient to learn major class and place features cross-lingually. △ Less

Submitted 12 June, 2025; v1 submitted 3 April, 2025; originally announced April 2025.

Comments: Accepted to CoNLL 2025

Rubrik's Cube: Testing a New Rubric for Evaluating Explanations on the CUBE dataset

Authors: Diana Galvan-Sosa, Gabrielle Gaudeau, Pride Kavumba, Yunmeng Li, Hongyi gu, Zheng Yuan, Keisuke Sakaguchi, Paula Buttery

Abstract: The performance and usability of Large-Language Models (LLMs) are driving their use in explanation generation tasks. However, despite their widespread adoption, LLM explanations have been found to be unreliable, making it difficult for users to distinguish good from bad explanations. To address this issue, we present Rubrik's CUBE, an education-inspired rubric and a dataset of 26k explanations, wr… ▽ More The performance and usability of Large-Language Models (LLMs) are driving their use in explanation generation tasks. However, despite their widespread adoption, LLM explanations have been found to be unreliable, making it difficult for users to distinguish good from bad explanations. To address this issue, we present Rubrik's CUBE, an education-inspired rubric and a dataset of 26k explanations, written and later quality-annotated using the rubric by both humans and six open- and closed-source LLMs. The CUBE dataset focuses on two reasoning and two language tasks, providing the necessary diversity for us to effectively test our proposed rubric. Using Rubrik, we find that explanations are influenced by both task and perceived difficulty. Low quality stems primarily from a lack of conciseness in LLM-generated explanations, rather than cohesion and word choice. The full dataset, rubric, and code are available at https://github.com/RubriksCube/rubriks_cube. △ Less

Submitted 4 June, 2025; v1 submitted 31 March, 2025; originally announced March 2025.

Comments: 10 main pages (24 appendix pages), 9 figures, accepted to ACL 2025

ACM Class: I.2.7

From Babble to Words: Pre-Training Language Models on Continuous Streams of Phonemes

Authors: Zébulon Goriely, Richard Diehl Martinez, Andrew Caines, Lisa Beinborn, Paula Buttery

Abstract: Language models are typically trained on large corpora of text in their default orthographic form. However, this is not the only option; representing data as streams of phonemes can offer unique advantages, from deeper insights into phonological language acquisition to improved performance on sound-based tasks. The challenge lies in evaluating the impact of phoneme-based training, as most benchmar… ▽ More Language models are typically trained on large corpora of text in their default orthographic form. However, this is not the only option; representing data as streams of phonemes can offer unique advantages, from deeper insights into phonological language acquisition to improved performance on sound-based tasks. The challenge lies in evaluating the impact of phoneme-based training, as most benchmarks are also orthographic. To address this, we develop a pipeline to convert text datasets into a continuous stream of phonemes. We apply this pipeline to the 100-million-word pre-training dataset from the BabyLM challenge, as well as to standard language and grammatical benchmarks, enabling us to pre-train and evaluate a model using phonemic input representations. Our results show that while phoneme-based training slightly reduces performance on traditional language understanding tasks, it offers valuable analytical and practical benefits. △ Less

Submitted 30 October, 2024; originally announced October 2024.

Less is More: Pre-Training Cross-Lingual Small-Scale Language Models with Cognitively-Plausible Curriculum Learning Strategies

Authors: Suchir Salhan, Richard Diehl Martinez, Zébulon Goriely, Paula Buttery

Abstract: Curriculum Learning has been a popular strategy to improve the cognitive plausibility of Small-Scale Language Models (SSLMs) in the BabyLM Challenge. However, it has not led to considerable improvements over non-curriculum models. We assess whether theoretical linguistic acquisition theories can be used to specify more fine-grained curriculum learning strategies, creating age-ordered corpora of Ch… ▽ More Curriculum Learning has been a popular strategy to improve the cognitive plausibility of Small-Scale Language Models (SSLMs) in the BabyLM Challenge. However, it has not led to considerable improvements over non-curriculum models. We assess whether theoretical linguistic acquisition theories can be used to specify more fine-grained curriculum learning strategies, creating age-ordered corpora of Child-Directed Speech for four typologically distant language families to implement SSLMs and acquisition-inspired curricula cross-lingually. Comparing the success of three objective curricula (Growing, Inwards and MMM) that precisely replicate the predictions of acquisition theories on a standard SSLM architecture, we find fine-grained acquisition-inspired curricula can outperform non-curriculum baselines and performance benefits of curricula strategies in SSLMs can be derived by specifying fine-grained language-specific curricula that precisely replicate language acquisition theories. △ Less

Submitted 21 February, 2025; v1 submitted 30 October, 2024; originally announced October 2024.

Comments: BabyLM Shared Task 2024 (Accepted, Poster), co-located in EMNLP 2024

Mitigating Frequency Bias and Anisotropy in Language Model Pre-Training with Syntactic Smoothing

Authors: Richard Diehl Martinez, Zebulon Goriely, Andrew Caines, Paula Buttery, Lisa Beinborn

Abstract: Language models strongly rely on frequency information because they maximize the likelihood of tokens during pre-training. As a consequence, language models tend to not generalize well to tokens that are seldom seen during training. Moreover, maximum likelihood training has been discovered to give rise to anisotropy: representations of tokens in a model tend to cluster tightly in a high-dimensiona… ▽ More Language models strongly rely on frequency information because they maximize the likelihood of tokens during pre-training. As a consequence, language models tend to not generalize well to tokens that are seldom seen during training. Moreover, maximum likelihood training has been discovered to give rise to anisotropy: representations of tokens in a model tend to cluster tightly in a high-dimensional cone, rather than spreading out over their representational capacity. Our work introduces a method for quantifying the frequency bias of a language model by assessing sentence-level perplexity with respect to token-level frequency. We then present a method for reducing the frequency bias of a language model by inducing a syntactic prior over token representations during pre-training. Our Syntactic Smoothing method adjusts the maximum likelihood objective function to distribute the learning signal to syntactically similar tokens. This approach results in better performance on infrequent English tokens and a decrease in anisotropy. We empirically show that the degree of anisotropy in a model correlates with its frequency bias. △ Less

Submitted 15 October, 2024; originally announced October 2024.

Tending Towards Stability: Convergence Challenges in Small Language Models

Authors: Richard Diehl Martinez, Pietro Lesci, Paula Buttery

Abstract: Increasing the number of parameters in language models is a common strategy to enhance their performance. However, smaller language models remain valuable due to their lower operational costs. Despite their advantages, smaller models frequently underperform compared to their larger counterparts, even when provided with equivalent data and computational resources. Specifically, their performance te… ▽ More Increasing the number of parameters in language models is a common strategy to enhance their performance. However, smaller language models remain valuable due to their lower operational costs. Despite their advantages, smaller models frequently underperform compared to their larger counterparts, even when provided with equivalent data and computational resources. Specifically, their performance tends to degrade in the late pretraining phase. This is anecdotally attributed to their reduced representational capacity. Yet, the exact causes of this performance degradation remain unclear. We use the Pythia model suite to analyse the training dynamics that underlie this phenomenon. Across different model sizes, we investigate the convergence of the Attention and MLP activations to their final state and examine how the effective rank of their parameters influences this process. We find that nearly all layers in larger models stabilise early in training - within the first 20% - whereas layers in smaller models exhibit slower and less stable convergence, especially when their parameters have lower effective rank. By linking the convergence of layers' activations to their parameters' effective rank, our analyses can guide future work to address inefficiencies in the learning dynamics of small models. △ Less

Submitted 15 October, 2024; originally announced October 2024.

Prompting open-source and commercial language models for grammatical error correction of English learner text

Authors: Christopher Davis, Andrew Caines, Øistein Andersen, Shiva Taslimipoor, Helen Yannakoudakis, Zheng Yuan, Christopher Bryant, Marek Rei, Paula Buttery

Abstract: Thanks to recent advances in generative AI, we are able to prompt large language models (LLMs) to produce texts which are fluent and grammatical. In addition, it has been shown that we can elicit attempts at grammatical error correction (GEC) from LLMs when prompted with ungrammatical input sentences. We evaluate how well LLMs can perform at GEC by measuring their performance on established benchm… ▽ More Thanks to recent advances in generative AI, we are able to prompt large language models (LLMs) to produce texts which are fluent and grammatical. In addition, it has been shown that we can elicit attempts at grammatical error correction (GEC) from LLMs when prompted with ungrammatical input sentences. We evaluate how well LLMs can perform at GEC by measuring their performance on established benchmark datasets. We go beyond previous studies, which only examined GPT* models on a selection of English GEC datasets, by evaluating seven open-source and three commercial LLMs on four established GEC benchmarks. We investigate model performance and report results against individual error types. Our results indicate that LLMs do not always outperform supervised English GEC models except in specific contexts -- namely commercial LLMs on benchmarks annotated with fluency corrections as opposed to minimal edits. We find that several open-source models outperform commercial ones on minimal edit benchmarks, and that in some settings zero-shot prompting is just as competitive as few-shot prompting. △ Less

Submitted 6 April, 2025; v1 submitted 15 January, 2024; originally announced January 2024.

Comments: 8 pages with appendices; accepted to ACL Findings 2024

CLIMB: Curriculum Learning for Infant-inspired Model Building

Authors: Richard Diehl Martinez, Zebulon Goriely, Hope McGovern, Christopher Davis, Andrew Caines, Paula Buttery, Lisa Beinborn

Abstract: We describe our team's contribution to the STRICT-SMALL track of the BabyLM Challenge. The challenge requires training a language model from scratch using only a relatively small training dataset of ten million words. We experiment with three variants of cognitively-motivated curriculum learning and analyze their effect on the performance of the model on linguistic evaluation tasks. In the vocabul… ▽ More We describe our team's contribution to the STRICT-SMALL track of the BabyLM Challenge. The challenge requires training a language model from scratch using only a relatively small training dataset of ten million words. We experiment with three variants of cognitively-motivated curriculum learning and analyze their effect on the performance of the model on linguistic evaluation tasks. In the vocabulary curriculum, we analyze methods for constraining the vocabulary in the early stages of training to simulate cognitively more plausible learning curves. In the data curriculum experiments, we vary the order of the training instances based on i) infant-inspired expectations and ii) the learning behavior of the model. In the objective curriculum, we explore different variations of combining the conventional masked language modeling task with a more coarse-grained word class prediction task to reinforce linguistic generalization capabilities. Our results did not yield consistent improvements over our own non-curriculum learning baseline across a range of linguistic benchmarks; however, we do find marginal gains on select tasks. Our analysis highlights key takeaways for specific combinations of tasks and settings which benefit from our proposed curricula. We moreover determine that careful selection of model architecture, and training hyper-parameters yield substantial improvements over the default baselines provided by the BabyLM challenge. △ Less

Submitted 15 November, 2023; originally announced November 2023.

On the application of Large Language Models for language teaching and assessment technology

Authors: Andrew Caines, Luca Benedetto, Shiva Taslimipoor, Christopher Davis, Yuan Gao, Oeistein Andersen, Zheng Yuan, Mark Elliott, Russell Moore, Christopher Bryant, Marek Rei, Helen Yannakoudakis, Andrew Mullooly, Diane Nicholls, Paula Buttery

Abstract: The recent release of very large language models such as PaLM and GPT-4 has made an unprecedented impact in the popular media and public consciousness, giving rise to a mixture of excitement and fear as to their capabilities and potential uses, and shining a light on natural language processing research which had not previously received so much attention. The developments offer great promise for e… ▽ More The recent release of very large language models such as PaLM and GPT-4 has made an unprecedented impact in the popular media and public consciousness, giving rise to a mixture of excitement and fear as to their capabilities and potential uses, and shining a light on natural language processing research which had not previously received so much attention. The developments offer great promise for education technology, and in this paper we look specifically at the potential for incorporating large language models in AI-driven language teaching and assessment systems. We consider several research areas and also discuss the risks and ethical considerations surrounding generative AI in education technology for language learners. Overall we find that larger language models offer improvements over previous models in text generation, opening up routes toward content generation which had not previously been plausible. For text generation they must be prompted carefully and their outputs may need to be reshaped before they are ready for use. For automated grading and grammatical error correction, tasks whose progress is checked on well-known benchmarks, early investigations indicate that large language models on their own do not improve on state-of-the-art results according to standard evaluation metrics. For grading it appears that linguistic features established in the literature should still be used for best performance, and for error correction it may be that the models can offer alternative feedback styles which are not measured sensitively with existing methods. In all cases, there is work to be done to experiment with the inclusion of large language models in education technology for language learners, in order to properly understand and report on their capacities and limitations, and to ensure that foreseeable risks such as misinformation and harmful bias are mitigated. △ Less

Submitted 17 July, 2023; originally announced July 2023.

Comments: Accepted at the AIED2023 workshop: Empowering Education with LLMs - the Next-Gen Interface and Content Generation

Probing for targeted syntactic knowledge through grammatical error detection

Authors: Christopher Davis, Christopher Bryant, Andrew Caines, Marek Rei, Paula Buttery

Abstract: Targeted studies testing knowledge of subject-verb agreement (SVA) indicate that pre-trained language models encode syntactic information. We assert that if models robustly encode subject-verb agreement, they should be able to identify when agreement is correct and when it is incorrect. To that end, we propose grammatical error detection as a diagnostic probe to evaluate token-level contextual rep… ▽ More Targeted studies testing knowledge of subject-verb agreement (SVA) indicate that pre-trained language models encode syntactic information. We assert that if models robustly encode subject-verb agreement, they should be able to identify when agreement is correct and when it is incorrect. To that end, we propose grammatical error detection as a diagnostic probe to evaluate token-level contextual representations for their knowledge of SVA. We evaluate contextual representations at each layer from five pre-trained English language models: BERT, XLNet, GPT-2, RoBERTa, and ELECTRA. We leverage public annotated training data from both English second language learners and Wikipedia edits, and report results on manually crafted stimuli for subject-verb agreement. We find that masked language models linearly encode information relevant to the detection of SVA errors, while the autoregressive models perform on par with our baseline. However, we also observe a divergence in performance when probes are trained on different training sets, and when they are evaluated on different syntactic constructions, suggesting the information pertaining to SVA error detection is not robustly encoded. △ Less

Submitted 28 October, 2022; originally announced October 2022.

Comments: CoNLL 2022

Constructing Open Cloze Tests Using Generation and Discrimination Capabilities of Transformers

Authors: Mariano Felice, Shiva Taslimipoor, Paula Buttery

Abstract: This paper presents the first multi-objective transformer model for constructing open cloze tests that exploits generation and discrimination capabilities to improve performance. Our model is further enhanced by tweaking its loss function and applying a post-processing re-ranking algorithm that improves overall test structure. Experiments using automatic and human evaluation show that our approach… ▽ More This paper presents the first multi-objective transformer model for constructing open cloze tests that exploits generation and discrimination capabilities to improve performance. Our model is further enhanced by tweaking its loss function and applying a post-processing re-ranking algorithm that improves overall test structure. Experiments using automatic and human evaluation show that our approach can achieve up to 82% accuracy according to experts, outperforming previous work and baselines. We also release a collection of high-quality open cloze tests along with sample system output and human annotations that can serve as a future benchmark. △ Less

Submitted 14 April, 2022; originally announced April 2022.

Comments: Accepted at Findings of ACL 2022

The Teacher-Student Chatroom Corpus

Authors: Andrew Caines, Helen Yannakoudakis, Helena Edmondson, Helen Allen, Pascual Pérez-Paredes, Bill Byrne, Paula Buttery

Abstract: The Teacher-Student Chatroom Corpus (TSCC) is a collection of written conversations captured during one-to-one lessons between teachers and learners of English. The lessons took place in an online chatroom and therefore involve more interactive, immediate and informal language than might be found in asynchronous exchanges such as email correspondence. The fact that the lessons were one-to-one mean… ▽ More The Teacher-Student Chatroom Corpus (TSCC) is a collection of written conversations captured during one-to-one lessons between teachers and learners of English. The lessons took place in an online chatroom and therefore involve more interactive, immediate and informal language than might be found in asynchronous exchanges such as email correspondence. The fact that the lessons were one-to-one means that the teacher was able to focus exclusively on the linguistic abilities and errors of the student, and to offer personalised exercises, scaffolding and correction. The TSCC contains more than one hundred lessons between two teachers and eight students, amounting to 13.5K conversational turns and 133K words: it is freely available for research use. We describe the corpus design, data collection procedure and annotations added to the text. We perform some preliminary descriptive analyses of the data and consider possible uses of the TSCC. △ Less

Submitted 13 November, 2020; originally announced November 2020.

Comments: NLP4CALL

Adaptive Forgetting Curves for Spaced Repetition Language Learning

Authors: Ahmed Zaidi, Andrew Caines, Russell Moore, Paula Buttery, Andrew Rice

Abstract: The forgetting curve has been extensively explored by psychologists, educationalists and cognitive scientists alike. In the context of Intelligent Tutoring Systems, modelling the forgetting curve for each user and knowledge component (e.g. vocabulary word) should enable us to develop optimal revision strategies that counteract memory decay and ensure long-term retention. In this study we explore a… ▽ More The forgetting curve has been extensively explored by psychologists, educationalists and cognitive scientists alike. In the context of Intelligent Tutoring Systems, modelling the forgetting curve for each user and knowledge component (e.g. vocabulary word) should enable us to develop optimal revision strategies that counteract memory decay and ensure long-term retention. In this study we explore a variety of forgetting curve models incorporating psychological and linguistic features, and we use these models to predict the probability of word recall by learners of English as a second language. We evaluate the impact of the models and their features using data from an online vocabulary teaching platform and find that word complexity is a highly informative feature which may be successfully learned by a neural network model. △ Less

Submitted 23 April, 2020; originally announced April 2020.

Comments: Artificial Intelligence for Education 2020 (AIED)