Conspiracy theories are anti-authoritarian narratives that lead to social conflict, impacting how people perceive political information. To help in understanding this issue, we introduce the Conspiracy Frame: a fine-grained semantic representation of conspiratorial narratives derived from frame-semantics and semiotics, which spawned the Conspiracy Frames (Con.Fra.) dataset: a corpus of Telegram messages annotated at span-level. The Conspiracy Frame and Con.Fra. dataset contribute to the implementation of a more generalizable understanding and recognition of conspiracy theories. We observe the ability of LLMs to recognize this phenomenon in-domain and out-of-domain, investigating the role that frames may have in supporting this task. Results show that, while the injection of frames in an in-context approach does not lead to clear increase of performance, it has potential; the mapping of annotated spans with FrameNet shows abstract semantic patterns (e.g., `Kinship', `Ingest_substance') that potentially pave the way for a more semantically- and semiotically-aware detection of conspiratorial narratives.

In High Energy Physics, as in many other fields of science, the application of machine learning techniques has been crucial in advancing our understanding of fundamental phenomena. Increasingly, deep learning models are applied to analyze both simulated and experimental data. In most experiments, a rigorous regime of testing for physically motivated systematic uncertainties is in place. The numerical evaluation of these tests for differences between the data on the one side and simulations on the other side quantifies the effect of potential sources of mismodelling on the machine learning output. In addition, thorough comparisons of marginal distributions and (linear) feature correlations between data and simulation in "control regions" are applied. However, the guidance by physical motivation, and the need to constrain comparisons to specific regions, does not guarantee that all possible sources of deviations have been accounted for. We therefore propose a new adversarial attack - the CONSERVAttack - designed to exploit the remaining space of hypothetical deviations between simulation and data after the above mentioned tests. The resulting adversarial perturbations are consistent within the uncertainty bounds - evading standard validation checks - while successfully fooling the underlying model. We further propose strategies to mitigate such vulnerabilities and argue that robustness to adversarial effects must be considered when interpreting results from deep learning in particle physics.

In this paper, we present a new algorithm, MiniFool, that implements physics-inspired adversarial attacks for testing neural network-based classification tasks in particle and astroparticle physics. While we initially developed the algorithm for the search for astrophysical tau neutrinos with the IceCube Neutrino Observatory, we apply it to further data from other science domains, thus demonstrating its general applicability. Here, we apply the algorithm to the well-known MNIST data set and furthermore, to public data from the CMS Experiment at the Large Hadron Collider. The algorithm is based on minimizing a cost function that combines a χ ^2 based test-statistic with the deviation from the desired target score. The test statistic quantifies the probability of the perturbations applied to the data based on the experimental uncertainties. For our studied use cases, we find that the likelihood of a flipped classification differs for both the initially correctly and incorrectly classified events. When testing changes of the classifications as a function of an attack parameter that scales the experimental uncertainties, the robustness of the network decision can be assessed. Furthermore, this allows testing the robustness of the classification of unlabeled experimental data.

Machine-learning models in high-energy physics are often trained on simulated data, where fully simulated samples are computationally expensive while fast simulation provides large statistics at reduced realism. In this work, we systematically study transfer learning between fast-simulated and fully simulated datasets in a realistic LHC environment. We consider three representative tasks, signal-background classification, quark-gluon jet tagging, and missing transverse energy reconstruction, using dense neural networks, graph neural networks, and transformer-based architectures. Models are pretrained on ATLAS-like fast simulation and adapted to CMS-like fast simulation and to fully simulated ATLAS Open Data. Across all tasks, pretrained models consistently outperform independently trained baselines and require significantly less target-domain training data, typically reducing the needed statistics by about a factor of two. These results demonstrate that fast simulation can be used to learn robust, reusable representations and motivate publishing trained models as reusable scientific assets beyond large foundation models.

Accurate Monte Carlo (MC) modelling in high-energy physics is challenging, particularly in complex scenarios where simulations fail to reproduce observed data. In practice, experimental information is often limited to one-dimensional (1D) distributions, while mismodelling arises in a multidimensional feature space. This restricts traditional correction methods, as one-dimensional reweighting ignores correlations and fully multidimensional approaches require large target datasets. We propose a neural network-based method that operates under these constraints by learning a transformation of simulated events that reproduces the available 1D target distributions while remaining close to the original simulation. This minimal-deviation principle preserves the global correlation structure of the baseline model while enabling targeted corrections of mismodelled features. Using controlled studies with simulated pseudo-data, we show that the method improves agreement with target distributions and maintains a consistent multidimensional structure. The approach is designed for complex, high-dimensional analyses where traditional techniques are insufficient, providing a scalable way to enhance MC modelling under limited information.

Pluralism, the capacity to engage with diverse perspectives without collapsing them into a single viewpoint, is critical for developing large language models that faithfully reflect human heterogeneity. Yet this characteristic has not been carefully examined within the LLM research community and remains absent from most alignment studies. Debate-oriented sources provide a natural entry point for pluralism research. Previous work builds on online debate sources but remains constrained by costly human validation. Other debate-rich platforms such as Reddit and Kialo also offer promising material: Reddit provides linguistic diversity and scale but lacks clear argumentative structure, while Kialo supplies explicit pro/con graphs but remains overly concise and detached from natural discourse. We introduce PERSPECTRA, a pluralist benchmark that integrates the structural clarity of Kialo debate graphs with the linguistic diversity of real Reddit discussions. Using a controlled retrieval-and-expansion pipeline, we construct 3,810 enriched arguments spanning 762 pro/con stances on 100 controversial topics. Each opinion is expanded into multiple naturalistic variants, enabling robust evaluation of pluralism. We initialise three tasks with PERSPECTRA: opinion counting (identifying distinct viewpoints), opinion matching (aligning supporting stances and discourse to source opinions), and polarity check (inferring aggregate stance in mixed discourse). Experiments with state-of-the-art open-source and proprietary LLMs, highlight systematic failures, such as overestimating the number of viewpoints and misclassifying concessive structures, underscoring the difficulty of pluralism-aware understanding and reasoning. By combining diversity with structure, PERSPECTRA establishes the first scalable, configurable benchmark for evaluating how well models represent, distinguish, and reason over multiple perspectives. We release PERSPECTRA as a resource with flexible configurations, enabling the creation of tasks beyond the demo tasks presented in this paper, and fostering progress toward pluralism-sensitive systems that more faithfully capture human heterogeneity.

The use of large language models (LLMs) such as ChatGPT (GPT-4/GPT-5) for mental health is growing rapidly, including their use to assess and help people with mood disorders, such as depression. However, we have a limited understanding of the schema of mental disorders in these language models, that is, how they internally associate and interpret the symptoms. In this work, we used contemporary measurement theory to decode how GPT-4 and GPT-5 interrelate depressive symptoms, providing an explanation of how large language models behave in clinical applications. First, we evaluated whether these models can reliably detect depressive constructs, finding that GPT-4 (a) demonstrated strong convergent validity with standard instruments and expert judgments (r = .70-.81). We mapped the model's internal schema of depression, showing that it (b) organized symptoms into an interconnected network (symptom correlations r = .23-.78) that largely aligned with the established clinical literature. However, it (c) underemphasized the relationship between suicidality and other symptoms while overemphasizing the importance of psychomotor symptoms. Finally, the associations (d) suggested novel hypotheses for symptom mechanisms, for instance, indicating that sleep and fatigue are broadly influenced by other depressive symptoms, whereas worthlessness/guilt is selectively associated with depressed mood. GPT-5 showed slightly lower convergence with self-report, which our analysis attributed to changes in the relationships among symptoms. These insights provide a generalizable, empirical foundation for understanding language models' mental health assessments across models and disorders. (PsycInfo Database Record (c) 2026 APA, all rights reserved).

Fine-tuning lets practitioners repurpose aligned large language models (LLMs) for new domains, yet recent work reveals emergent misalignment (EMA): Even a small, domain-specific fine-tune can induce harmful behaviors far outside the target domain. Even in the case where model weights are hidden behind a fine-tuning API, this gives attackers inadvertent access to a broadly misaligned model in a way that can be hard to detect from the fine-tuning data alone. We present the first systematic study of in-training safeguards against EMA that are practical for providers who expose fine-tuning via an API: We evaluate whether they a) prevent broad misalignment, b) allow narrow misalignment, c) learn well on benign tasks, and d) remain coherent. We investigate four training regularization interventions: (i) KL-divergence regularization toward a safe reference model, (ii) 𝓁_2 distance in feature space, (iii) preventative steering with an evil persona vector, and (iv) interleaving training examples from a general instruct-tuning dataset. We demonstrate that selecting interleaving data by the perplexity gap between aligned and misaligned models yields the best results overall.

We propose an agentic data augmentation method for Aspect-Based Sentiment Analysis (ABSA) that uses iterative generation and verification to produce high quality synthetic training examples. To isolate the effect of agentic structure, we also develop a closely matched prompting-based baseline using the same model and instructions. Both methods are evaluated across three ABSA subtasks (Aspect Term Extraction (ATE), Aspect Sentiment Classification (ATSC), and Aspect Sentiment Pair Extraction (ASPE)), four SemEval datasets, and two encoder-decoder models: T5-Base and Tk-Instruct. Our results show that the agentic augmentation outperforms raw prompting in label preservation of the augmented data, especially when the tasks require aspect term generation. In addition, when combined with real data, agentic augmentation provides higher gains, consistently outperforming prompting-based generation. These benefits are most pronounced for T5-Base, while the more heavily pretrained Tk-Instruct exhibits smaller improvements. As a result, augmented data helps T5-Base achieve comparable performance with its counterpart.

This paper systematically examines cross-lingual disparities in pre-training source and factuality alignment in Large Language Model (LLM) answers for multilingual healthcare Q A across English, German, Turkish, Chinese (Mandarin), and Italian. To support this analysis, we (i) constructed MultiWikiHealthCare, a multilingual dataset derived from Wikipedia; (ii) used it to examine cross-lingual differences in healthcare-related coverage; (iii) evaluated the alignment between LLM-generated responses and these reference sources; and (iv) conducted a case study on factual alignment through the use of contextual information and Retrieval-Augmented Generation (RAG). Our findings reveal substantial cross-lingual disparities in both Wikipedia coverage and LLM factual alignment. Based on our Wikipedia analysis, the lowest alignment is observed between English and Chinese Wikipedia pages. Across models, responses align more with English Wikipedia, even when the prompts are non-English. We further show that providing contextual excerpts from non-English Wikipedia at inference time effectively shifts factual alignment toward target knowledge.

We present an overview of the Colliding With Adversaries challenge on robust learning in High Energy Physics, held at ECML PKDD 2025. The challenge was split into two tasks: (1) generating adversarial examples to attack a deep learning model, and (2) developing a model robust to unseen adversarial attacks while maintaining strong performance on clean data. Both tasks used a tabular dataset derived from two simulated CMS LHC collision processes, featuring 87 continuous features per sample. We summarize the challenge setup, evaluation metrics, dataset characteristics, and submission statistics. Finally, we describe the methods of the winning team, Team APTH, who employed a white-box gradient-based attack for the Adversarial Attack Task and a method-aware robustness strategy using an ensemble of feature-embedding models for the Robust Model Task. The challenge highlights the potential of adversarial deep learning in high-energy physics, promotes research on robust models for tabular data, and seeks to introduce non-physicists to the rich datasets typical of high-energy physics experiments.

Due to the scarcity of labeled dialectal speech, audio dialect classification is a challenging task for most languages, including Swiss German. In this work, we explore the ability of large language models (LLMs) as agents in understanding the dialects and whether they can show comparable performance to models such as HuBERT in dialect classification. In addition, we provide an LLM baseline and a human linguist one. Our approach uses phonetic transcriptions produced by ASR systems and combines them with linguistic resources such as dialect feature maps, vowel history, and rules. Our findings indicate that, when linguistic information is provided, the LLM predictions improve. The human baseline shows that automatically generated transcriptions can be beneficial for such classifications, but also presents opportunities for improvement.

The EmpathiCH workshop series has, over three iterations, unpacked how empathy is conceptualized, measured, and used in HCI, identifying both its potential benefits and notable pitfalls. Despite these discussions, the diverse roles of empathy in research and practice remain fragmented and under-theorized. This fourth iteration seeks to consolidate perspectives by situating empathy within a sociomaterial framework. We propose exploring three dimensions—technology, social practices, and context—that shape how empathy is conceptualized and applied. The workshop will combine an interactive, discussion-centric format enabling participants to share experiences, debate perspectives, and collaboratively analyze cases across these dimensions. Outcomes will contribute to co-developing a sociomaterial taxonomy for empathy in HCI, offering conceptual clarity and practical guidance for design. Participants will engage in critical dialogue, connect with peers, and contribute directly to shaping the future of empathy-centered approaches in HCI.

Neural networks (NNs) are inherently multidimensional classifiers that learn complex, non-linear relationships among input observables. While their flexibility enables unprecedented performance in high-energy physics (HEP) analyses, it also makes them sensitive to small variations in their inputs. Consequently, the propagation and estimation of systematic uncertainties in NN-based models remain an open challenge. There are indications that uncertainties derived in control regions or from nominal variations of input features can underestimate the true model uncertainty, potentially leaving biases unaccounted for. Inspired by insights from adversarial-attack studies in machine learning, we explore how subtle perturbations, fully consistent with the experimental uncertainties on the input observables, can lead to substantial changes in NN outputs, while keeping the one-dimensional and correlated input distributions nearly unchanged. Using a set of representative HEP tasks, including event classification and object identification, and testing across a variety of network architectures, we demonstrate that networks can be systematically "fooled" at significant rates within the allowed uncertainty envelopes. Building on this observation, we introduce a quantitative framework to probe and measure the hidden sensitivity of neural networks to realistic experimental variations, providing a practical path to evaluate and control their systematic uncertainty in physics analyses.

When LLM agents work together, they seem to be more powerful than a single LLM in mathematical question answering. However, are they also more robust to adversarial inputs? We investigate this question using adversarially perturbed math questions. These perturbations include punctuation noise with three intensities (10

Reasoning in large language models is often discussed as a single capability, but some of its gains may stem from simpler underlying operations. We examine two such primitives, recall and state-tracking, through five controlled task families centered on state-based recall, and compare matched transformer and hybrid architectures with and without reasoning augmentation. Across the suite, reasoning-augmented variants substantially outperform instruction-only variants, often by large margins. This pattern is consistent with the State over Tokens view: externalized reasoning traces help because they carry the intermediate state forward in token space. By contrast, hybrid inductive bias does not yield a uniform advantage in accuracy once reasoning tokens are available. When architectural differences do appear, they follow task structure: the hybrid Think model is more robust on strictly sequential chained updates, whereas the transformer Think model is more robust on flat multi-hop retrieval. We therefore cast the main contribution of this study as a descriptive account of what drives performance on state-based recall tasks: reasoning-token augmentation appears to be the dominant factor, while hybrid advantages are narrower, task-dependent, and potentially more about inference efficiency than overall capability. We also release the codebase and data required to reproduce these results.

Social interactions are shaped by the interplay of dispositional traits and situational context, yet systematically investigating how personality configurations between individuals jointly influence social behavior across diverse social contexts remains methodologically challenging. We address this gap by introducing a simulation pipeline adapted from the CHARISMA framework, which employs well-known movie characters and public figures as psychologically grounded agents for multi-LLM social simulation using a method we term personality anchoring. We present a large-scale empirical study examining how dyadic Agreeableness composition influences social interaction outcomes across 1,010 simulated conversations. Our results reveal a monotonic relationship between dyadic Agreeableness composition and shared goal achievement, with Homogeneous-Agreeable pairs achieving success 10 times the rate of Homogeneous-Disagreeable pairs (62

Theory of Mind (ToM) refers to an agent's ability to model the internal states of others. Contributing to the debate whether large language models (LLMs) exhibit genuine ToM capabilities, our study investigates their ToM robustness using perturbations on false-belief tasks and examines the potential of Chain-of-Thought prompting (CoT) to enhance performance and explain the LLM's decision. We introduce a handcrafted, richly annotated ToM dataset, including classic and perturbed false belief tasks, the corresponding spaces of valid reasoning chains for correct task completion, subsequent reasoning faithfulness, task solutions, and propose metrics to evaluate reasoning chain correctness and to what extent final answers are faithful to reasoning traces of the generated CoT. We show a steep drop in ToM capabilities under task perturbation for all evaluated LLMs, questioning the notion of any robust form of ToM being present. While CoT prompting improves the ToM performance overall in a faithful manner, it surprisingly degrades accuracy for some perturbation classes, indicating that selective application is necessary.