Haim Dubossarsky

Joe Stacey, Pasquale Minervini, Haim Dubossarsky et al.

Current Natural Language Inference (NLI) models achieve impressive results, sometimes outperforming humans when evaluating on in-distribution test sets. However, as these models are known to learn from annotation artefacts and dataset biases, it is unclear to what extent the models are learning the task of NLI instead of learning from shallow heuristics in their training data. We address this issue by introducing a logical reasoning framework for NLI, creating highly transparent model decisions that are based on logical rules. Unlike prior work, we show that improved interpretability can be achieved without decreasing the predictive accuracy. We almost fully retain performance on SNLI, while also identifying the exact hypothesis spans that are responsible for each model prediction. Using the e-SNLI human explanations, we verify that our model makes sensible decisions at a span level, despite not using any span labels during training. We can further improve model performance and span-level decisions by using the e-SNLI explanations during training. Finally, our model is more robust in a reduced data setting. When training with only 1,000 examples, out-of-distribution performance improves on the MNLI matched and mismatched validation sets by 13% and 16% relative to the baseline. Training with fewer observations yields further improvements, both in-distribution and out-of-distribution.

Nina Tahmasebi, Haim Dubossarsky

In this chapter we provide an overview of computational modeling for semantic change using large and semi-large textual corpora. We aim to provide a key for the interpretation of relevant methods and evaluation techniques, and also provide insights into important aspects of the computational study of semantic change. We discuss the pros and cons of different classes of models with respect to the properties of the data from which one wishes to model semantic change, and which avenues are available to evaluate the results.

Aideen Fay, Inés García-Redondo, Qiquan Wang et al.

Existing interpretability methods for Large Language Models (LLMs) predominantly capture linear directions or isolated features. This overlooks the high-dimensional, relational, and nonlinear geometry of model representations. We apply persistent homology (PH) to characterize how adversarial inputs reshape the geometry and topology of internal representation spaces of LLMs. This phenomenon, especially when considered across operationally different attack modes, remains poorly understood. We analyze six models (3.8B to 70B parameters) under two distinct attacks, indirect prompt injection and backdoor fine--tuning, and show that a consistent topological signature persists throughout. Adversarial inputs induce topological compression, where the latent space becomes structurally simpler, collapsing the latent space from varied, compact, small-scale features into fewer, dominant, large-scale ones. This signature is architecture-agnostic, emerges early in the network, and is highly discriminative across layers. By quantifying the shape of activation point clouds and neuron-level information flow, our framework reveals geometric invariants of representational change that complement existing linear interpretability methods.

Roksana Goworek, Haim Dubossarsky

Lexical semantic change detection (LSCD) increasingly relies on contextualised language model embeddings, yet most approaches still quantify change using a small set of semantic change metrics, primarily Average Pairwise Distance (APD) and cosine distance over word prototypes (PRT). We introduce Average Minimum Distance (AMD) and Symmetric Average Minimum Distance (SAMD), new measures that quantify semantic change via local correspondence between word usages across time periods. Across multiple languages, encoder models, and representation spaces, we show that AMD often provides more robust performance, particularly under dimensionality reduction and with non-specialised encoders, while SAMD excels with specialised encoders. We suggest that LSCD may benefit from considering alternative semantic change metrics beyond APD and PRT, with AMD offering a robust option for contextualised embedding-based analysis.

Yuxuan Liu, Haim Dubossarsky, Ruth Ahnert

This paper examines how science fiction destabilises ontological categories by measuring conceptual permeability across the terms human, animal, and machine using masked language modelling (MLM). Drawing on corpora of science fiction (Gollancz SF Masterworks) and general fiction (NovelTM), we operationalise Darko Suvin's theory of estrangement as computationally measurable deviation in token prediction, using RoBERTa to generate lexical substitutes for masked referents and classifying them via Gemini. We quantify conceptual slippage through three metrics: retention rate, replacement rate, and entropy, mapping the stability or disruption of category boundaries across genres. Our findings reveal that science fiction exhibits heightened conceptual permeability, particularly around machine referents, which show significant cross-category substitution and dispersion. Human terms, by contrast, maintain semantic coherence and often anchor substitutional hierarchies. These patterns suggest a genre-specific restructuring within anthropocentric logics. We argue that estrangement in science fiction operates as a controlled perturbation of semantic norms, detectable through probabilistic modelling, and that MLMs, when used critically, serve as interpretive instruments capable of surfacing genre-conditioned ontological assumptions. This study contributes to the methodological repertoire of computational literary studies and offers new insights into the linguistic infrastructure of science fiction.

Francesco Periti, Pierluigi Cassotti, Haim Dubossarsky et al.

Modern language models are capable of contextualizing words based on their surrounding context. However, this capability is often compromised due to semantic change that leads to words being used in new, unexpected contexts not encountered during pre-training. In this paper, we model \textit{semantic change} by studying the effect of unexpected contexts introduced by \textit{lexical replacements}. We propose a \textit{replacement schema} where a target word is substituted with lexical replacements of varying relatedness, thus simulating different kinds of semantic change. Furthermore, we leverage the replacement schema as a basis for a novel \textit{interpretable} model for semantic change. We are also the first to evaluate the use of LLaMa for semantic change detection.

Naomi Baes, Raphaël Merx, Nick Haslam et al.

Lexical Semantic Change (LSC) provides insight into cultural and social dynamics. Yet, the validity of methods for measuring different kinds of LSC remains unestablished due to the absence of historical benchmark datasets. To address this gap, we propose LSC-Eval, a novel three-stage general-purpose evaluation framework to: (1) develop a scalable methodology for generating synthetic datasets that simulate theory-driven LSC using In-Context Learning and a lexical database; (2) use these datasets to evaluate the sensitivity of computational methods to synthetic change; and (3) assess their suitability for detecting change in specific dimensions and domains. We apply LSC-Eval to simulate changes along the Sentiment, Intensity, and Breadth (SIB) dimensions, as defined in the SIBling framework, using examples from psychology. We then evaluate the ability of selected methods to detect these controlled interventions. Our findings validate the use of synthetic benchmarks, demonstrate that tailored methods effectively detect changes along SIB dimensions, and reveal that a state-of-the-art LSC model faces challenges in detecting affective dimensions of LSC. LSC-Eval offers a valuable tool for dimension- and domain-specific benchmarking of LSC methods, with particular relevance to the social sciences.

Roksana Goworek, Harpal Karlcut, Muhammad Shezad et al.

This paper addresses the critical need for high-quality evaluation datasets in low-resource languages to advance cross-lingual transfer. While cross-lingual transfer offers a key strategy for leveraging multilingual pretraining to expand language technologies to understudied and typologically diverse languages, its effectiveness is dependent on quality and suitable benchmarks. We release new sense-annotated datasets of sentences containing polysemous words, spanning ten low-resource languages across diverse language families and scripts. To facilitate dataset creation, the paper presents a demonstrably beneficial semi-automatic annotation method. The utility of the datasets is demonstrated through Word-in-Context (WiC) formatted experiments that evaluate transfer on these low-resource languages. Results highlight the importance of targeted dataset creation and evaluation for effective polysemy disambiguation in low-resource settings and transfer studies. The released datasets and code aim to support further research into fair, robust, and truly multilingual NLP.

Roksana Goworek, Haim Dubossarsky

Cross-lingual transfer is central to modern NLP, enabling models to perform tasks in languages different from those they were trained on. A common assumption is that training on more languages improves zero-shot transfer. We test this on sense-aware tasks-polysemy and lexical semantic change-and find that multilinguality is not necessary for effective transfer. Our large-scale analysis across 28 languages reveals that other factors, such as differences in pretraining and fine-tuning data and evaluation artifacts, better explain the perceived benefits of multilinguality. We also release fine-tuned models and provide empirical baselines to support future research. While focused on two sense-aware tasks, our findings offer broader insights into cross-lingual transfer, especially for low-resource languages.

Francesco Periti, Haim Dubossarsky, Nina Tahmasebi

In the universe of Natural Language Processing, Transformer-based language models like BERT and (Chat)GPT have emerged as lexical superheroes with great power to solve open research problems. In this paper, we specifically focus on the temporal problem of semantic change, and evaluate their ability to solve two diachronic extensions of the Word-in-Context (WiC) task: TempoWiC and HistoWiC. In particular, we investigate the potential of a novel, off-the-shelf technology like ChatGPT (and GPT) 3.5 compared to BERT, which represents a family of models that currently stand as the state-of-the-art for modeling semantic change. Our experiments represent the first attempt to assess the use of (Chat)GPT for studying semantic change. Our results indicate that ChatGPT performs significantly worse than the foundational GPT version. Furthermore, our results demonstrate that (Chat)GPT achieves slightly lower performance than BERT in detecting long-term changes but performs significantly worse in detecting short-term changes.

Joe Stacey, Pasquale Minervini, Haim Dubossarsky et al.

With recent advances, neural models can achieve human-level performance on various natural language tasks. However, there are no guarantees that any explanations from these models are faithful, i.e. that they reflect the inner workings of the model. Atomic inference overcomes this issue, providing interpretable and faithful model decisions. This approach involves making predictions for different components (or atoms) of an instance, before using interpretable and deterministic rules to derive the overall prediction based on the individual atom-level predictions. We investigate the effectiveness of using LLM-generated facts as atoms, decomposing Natural Language Inference premises into lists of facts. While directly using generated facts in atomic inference systems can result in worse performance, with 1) a multi-stage fact generation process, and 2) a training regime that incorporates the facts, our fact-based method outperforms other approaches.

Dominik Schlechtweg, Nina Tahmasebi, Simon Hengchen et al.

Word meaning is notoriously difficult to capture, both synchronically and diachronically. In this paper, we describe the creation of the largest resource of graded contextualized, diachronic word meaning annotation in four different languages, based on 100,000 human semantic proximity judgments. We thoroughly describe the multi-round incremental annotation process, the choice for a clustering algorithm to group usages into senses, and possible - diachronic and synchronic - uses for this dataset.

Simon Hengchen, Nina Tahmasebi, Dominik Schlechtweg et al.

The computational study of lexical semantic change (LSC) has taken off in the past few years and we are seeing increasing interest in the field, from both computational sciences and linguistics. Most of the research so far has focused on methods for modelling and detecting semantic change using large diachronic textual data, with the majority of the approaches employing neural embeddings. While methods that offer easy modelling of diachronic text are one of the main reasons for the spiking interest in LSC, neural models leave many aspects of the problem unsolved. The field has several open and complex challenges. In this chapter, we aim to describe the most important of these challenges and outline future directions.

Dominik Schlechtweg, Barbara McGillivray, Simon Hengchen et al.

Lexical Semantic Change detection, i.e., the task of identifying words that change meaning over time, is a very active research area, with applications in NLP, lexicography, and linguistics. Evaluation is currently the most pressing problem in Lexical Semantic Change detection, as no gold standards are available to the community, which hinders progress. We present the results of the first shared task that addresses this gap by providing researchers with an evaluation framework and manually annotated, high-quality datasets for English, German, Latin, and Swedish. 33 teams submitted 186 systems, which were evaluated on two subtasks.

Joe Stacey, Pasquale Minervini, Haim Dubossarsky et al.

Natural Language Inference (NLI) datasets contain annotation artefacts resulting in spurious correlations between the natural language utterances and their respective entailment classes. These artefacts are exploited by neural networks even when only considering the hypothesis and ignoring the premise, leading to unwanted biases. Belinkov et al. (2019b) proposed tackling this problem via adversarial training, but this can lead to learned sentence representations that still suffer from the same biases. We show that the bias can be reduced in the sentence representations by using an ensemble of adversaries, encouraging the model to jointly decrease the accuracy of these different adversaries while fitting the data. This approach produces more robust NLI models, outperforming previous de-biasing efforts when generalised to 12 other datasets (Belinkov et al., 2019a; Mahabadi et al., 2020). In addition, we find that the optimal number of adversarial classifiers depends on the dimensionality of the sentence representations, with larger sentence representations being more difficult to de-bias while benefiting from using a greater number of adversaries.

Haim Dubossarsky, Ivan Vulić, Roi Reichart et al.

Performance in cross-lingual NLP tasks is impacted by the (dis)similarity of languages at hand: e.g., previous work has suggested there is a connection between the expected success of bilingual lexicon induction (BLI) and the assumption of (approximate) isomorphism between monolingual embedding spaces. In this work we present a large-scale study focused on the correlations between monolingual embedding space similarity and task performance, covering thousands of language pairs and four different tasks: BLI, parsing, POS tagging and MT. We hypothesize that statistics of the spectrum of each monolingual embedding space indicate how well they can be aligned. We then introduce several isomorphism measures between two embedding spaces, based on the relevant statistics of their individual spectra. We empirically show that 1) language similarity scores derived from such spectral isomorphism measures are strongly associated with performance observed in different cross-lingual tasks, and 2) our spectral-based measures consistently outperform previous standard isomorphism measures, while being computationally more tractable and easier to interpret. Finally, our measures capture complementary information to typologically driven language distance measures, and the combination of measures from the two families yields even higher task performance correlations.

Haim Dubossarsky, Simon Hengchen, Nina Tahmasebi et al.

State-of-the-art models of lexical semantic change detection suffer from noise stemming from vector space alignment. We have empirically tested the Temporal Referencing method for lexical semantic change and show that, by avoiding alignment, it is less affected by this noise. We show that, trained on a diachronic corpus, the skip-gram with negative sampling architecture with temporal referencing outperforms alignment models on a synthetic task as well as a manual testset. We introduce a principled way to simulate lexical semantic change and systematically control for possible biases.