Large language models typically undergo post-training to align them with safety policies but there exist many sophisticated jailbreaks that sidestep established safeguards. For instance, prior work by Andriushchenko et al. (2025) has found that changing the grammatical tense from present to past can be enough to elicit harmful responses. In this work, we uncover a more general failure of non-imperative syntactic forms. We demonstrate that this syntactic vulnerability exists in 16 models up to 70B parameters, using behavioral evaluation. To investigate the root cause, we apply causal mediation analysis, finding that refusal is partially conditioned on upstream syntactic features. By steering these purely syntactic features we are able to trigger and suppress refusal. Finally, we trace this ill-conditioning to linguistically biased post-training data of open-source models and show that increasing syntactic diversity can mitigate the issue. Our findings suggest that current alignment approaches introduce confounders that prevent a pure semantic grounding of the refusal decision.

Music festival lineups emerge from complex relationships among artists, genres, releases, labels, and past performances, making the prediction of future lineups a natural fit for temporal knowledge graph (TKG) forecasting. In this work, we present a TKG covering 380 festivals over 55 years, comprising more than 90K festival performance quadruples along with information on festivals, artist tours, and artist metadata, and release it as a resource for TKG forecasting evaluation. We formalize festival lineup forecasting as temporal link prediction between artists and festivals at future timestamps. We evaluate six TKG forecasting models on this task, analyze their capabilities and limitations, and compare them against Large Language Models applied zero-shot. Our resource complements existing TKG benchmarks by grounding evaluation in a concrete, real-world application domain.

Argumentation is an important means of communication. For describing especially arguments about consequences, the notion of effect relations has been introduced recently. We propose a method to extract effect relations from large text resources and apply it on encyclopedic and argumentative texts. By connecting the extracted relations, we generate a knowledge graph which we call effect graph. For evaluating the effect graph, we perform crowd and expert annotations and create a novel dataset. We demonstrate a possible use case of the effect graph by proposing a method for explaining arguments from consequences.

In this paper, we introduce the task of political coalition signal prediction from text, that is, the task of recognizing from the news coverage leading up to an election the (un)willingness of political parties to form a government coalition. We decompose our problem into two related, but distinct tasks: (i) predicting whether a reported statement from a politician or a journalist refers to a potential coalition and (ii) predicting the polarity of the signal – namely, whether the speaker is in favour of or against the coalition. For this, we explore the benefits of multi-task learning and investigate which setup and task formulation is best suited for each sub-task. We evaluate our approach, based on hand-coded newspaper articles, covering elections in three countries (Ireland, Germany, Austria) and two languages (English, German). Our results show that the multi-task learning approach can further improve results over a strong monolingual transfer learning baseline.

Operationalizing morality is crucial for understanding multiple aspects of society that have moral values at their core – such as riots, mobilizing movements, public debates, etc. Moral Foundations Theory (MFT) has become one of the most adopted theories of morality partly due to its accompanying lexicon, the Moral Foundation Dictionary (MFD), which offers a base for computationally dealing with morality. In this work, we exploit the MFD in a novel direction by investigating how well moral values are captured by KGs. We explore three widely used KGs, and provide concept-level analogues for the MFD. Furthermore, we propose several Personalized PageRank variations in order to score all the concepts and entities in the KGs with respect to their relevance to the different moral values. Our promising results help to progress the operationalization of morality in both NLP and KG communities.

In this paper, we perform modality prediction in financial dialogue. To this end, we introduce a new dataset and develop a binary classifier to detect strong or weak modal answers depending on surface, lexical, and semantic representations of the preceding question and financial features. To do so, we contrast different algorithms, feature categories, and fusion methods. Perhaps counter-intuitively, our results indicate that the strongest features for the given task are financial uncertainty measures such as market and individual firm risk.

Social media platforms have become an essential venue for online deliberation where users discuss arguments, debate, and form opinions. In this paper, we propose an unsupervised method to detect the stance of argumentative claims with respect to a topic. Most related work focuses on topic-specific supervised models that need to be trained for every emergent debate topic. To address this limitation, we propose a topic independent approach that focuses on a frequently encountered class of arguments, specifically, on arguments from consequences. We do this by extracting the effects that claims refer to, and proposing a means for inferring if the effect is a good or bad consequence. Our experiments provide promising results that are comparable to, and in particular regards even outperform BERT. Furthermore, we publish a novel dataset of arguments relating to consequences, annotated with Amazon Mechanical Turk.

Sentiment and stance are two important concepts for the analysis of arguments. We propose to add another perspective to the analysis, namely moral sentiment. We argue that moral values are crucial for ideological debates and can thus add useful information for argument mining. In the paper, we present different models for automatically predicting moral sentiment in debates and evaluate them on a manually annotated testset. We then apply our models to investigate how moral values in arguments relate to argument quality, stance and audience reactions.

We propose an unsupervised approach for assessing conceptual complexity of texts, based on spreading activation. Using DBpedia knowledge graph as a proxy to long-term memory, mentioned concepts become activated and trigger further activation as the text is sequentially traversed. Drawing inspiration from psycholinguistic theories of reading comprehension, we model memory processes such as semantic priming, sentence wrap-up, and forgetting. We show that our models capture various aspects of conceptual text complexity and significantly outperform current state of the art.

In this paper, we use NLP techniques to detect linguistic uncertainty in financial disclosures. Leveraging general-domain and domain-specific word embedding models, we automatically expand an existing dictionary of uncertainty triggers. We furthermore examine how an expert filtering affects the quality of such an expansion. We show that the dictionary expansions significantly improve regressions on stock return volatility. Lastly, we prove that the expansions significantly boost the automatic detection of uncertain sentences.

We provide several methods for sentence-alignment of texts with different complexity levels. Using the best of them, we sentence-align the Newsela corpora, thus providing large training materials for automatic text simplification (ATS) systems. We show that using this dataset, even the standard phrase-based statistical machine translation models for ATS can outperform the state-of-the-art ATS systems.