Current LLM pipelines account for only one possible tokenization for a given string, ignoring exponentially many alternative tokenizations during training and inference. For example, the standard Llama3 tokenization of penguin is [p,enguin], yet [peng,uin] is another perfectly valid alternative. In this paper, we show that despite LLMs being trained solely on one tokenization, they still retain semantic understanding of other tokenizations, raising questions about their implications in LLM safety. Put succinctly, we answer the following question: can we adversarially tokenize an obviously malicious string to evade safety and alignment restrictions? We show that not only is adversarial tokenization an effective yet previously neglected axis of attack, but it is also competitive against existing state-of-the-art adversarial approaches without changing the text of the harmful request. We empirically validate this exploit across three state-of-the-art LLMs and adversarial datasets, revealing a previously unknown vulnerability in subword models.

Geographic distribution shift arises when the distribution of locations on Earth in a training dataset is different from what is seen at test time. The most common approaches to tackling geographic distribution shift treat regions delimited by administrative boundaries such as countries or continents as separate domains and apply standard domain adaptation methods, ignoring geographic coordinates that are often available as metadata. This paper proposes the use of location encoders for training models that are more robust to geographic distribution shift. We show how both simple sine-cosine encoders and pre-trained location encoders can be used to improve standard domain adaptation methods for the special case of geographic distribution shift. Our proposed methods achieve state-of-the-art results on geo-tagged imagery datasets from the WILDS benchmark.

Longitudinal network data are essential for analyzing political, economic, and social systems and processes. In political science, these datasets are often generated through human annotation or supervised machine learning applied to evolving corpora. However, as semantic contexts shift over time, inferring dynamic interaction types on emerging issues among a diverse set of entities poses significant challenges, particularly in maintaining timely and consistent annotations. This paper presents the Expert-Augmented LLM Annotation (EALA) approach, which leverages Large Language Models (LLMs) in combination with historically annotated data and expert-constructed codebooks to extrapolate and extend datasets into future periods. We evaluate the performance and reliability of EALA using a dataset of climate negotiations. Our findings demonstrate that EALA effectively predicts nuanced interactions between negotiation parties and captures the evolution of topics over time. At the same time, we identify several limitations inherent to LLM-based annotation, highlighting areas for further improvement. Given the wide availability of codebooks and annotated datasets, EALA holds substantial promise for advancing research in political science and beyond.

Hypertensive retinopathy (HR) is a severe eye disease that may cause permanent vision loss if not diagnosed early. Traditional diagnostic methods are time-consuming and subjective, highlighting the need for an automated, reliable system. Existing studies often use a single Deep Learning (DL) model, struggling to distinguish HR stages. This study introduces a three-stage approach to enhance HR diagnosis accuracy. Initially, 14 CNN models were tested, identifying DenseNet169, MobileNet, and ResNet152 as the most effective. DenseNet169 achieved 87.73% accuracy, 87.75% precision, 87.73% recall, 87.67% F1-score, and 0.8359 Cohen's Kappa. MobileNet followed with 86.40% accuracy, 86.60% precision, 86.40% recall, 86.31% F1-score, and 0.8180 Cohen's Kappa. ResNet152 ranked third with 85.87% accuracy, 86.01% precision, 85.87% recall, 85.83% F1-score, and 0.8188 Cohen's Kappa. In the second stage, deep features from these models were fused and classified using Machine Learning (ML) algorithms (SVM, RF, XGBoost). SVM (sigmoid kernel) performed best with 92.00% accuracy, 91.93% precision, 92.00% recall, 91.91% F1-score, and 0.8930 Cohen's Kappa. The third stage applied meta-heuristic optimization (GA, ABC, PSO, HHO) for feature selection. HHO yielded 94.66% accuracy, precision, and recall, 94.64% F1-score, and 0.9286 Cohen's Kappa. The proposed approach surpassed single CNN models and previous studies in HR diagnosis accuracy and generalization.

Node classification is a key task in temporal graph learning (TGL). Real-life temporal graphs often introduce new node classes over time, but existing TGL methods assume a fixed set of classes. This assumption brings limitations, as updating models with full data is costly, while focusing only on new classes results in forgetting old ones. Graph continual learning (GCL) methods mitigate forgetting using old-class subsets but fail to account for their evolution. We define this novel problem as temporal graph continual learning (TGCL), which focuses on efficiently maintaining up-to-date knowledge of old classes. To tackle TGCL, we propose a selective learning framework that substitutes the old-class data with its subsets, Learning Towards the Future (LTF). We derive an upper bound on the error caused by such replacement and transform it into objectives for selecting and learning subsets that minimize classification error while preserving the distribution of the full old-class data. Experiments on three real-world datasets validate the effectiveness of LTF on TGCL.

Multiple-choice exam questions with "None of the above" (NA) options have been extensively studied in educational testing, in which existing research suggests that they better assess true knowledge. However, their impact on Large Language Models (LLMs) evaluation remains underexplored. Through systematic experiments with 28 LLMs on the MMLU benchmark, we examine how NA options affect model performance and confidence calibration. Our analysis reveals that NA options, when used as the correct answer, lead to a consistent 30-50\% performance drop across models regardless of scale--suggesting that LLMs lack the meta-cognitive ability to systematically evaluate and reject all given options when none are correct. This degradation shows strong domain dependence, with minimal impact on mathematical reasoning (14.6\% drop) but severe effects on tasks requiring uncertainty handling like business ethics (48.1\% drop). Our results highlight important implications for benchmark design and raise questions about LLMs' ability to handle uncertainty in real-world applications.

The KM3NeT Collaboration is installing the ARCA and ORCA neutrino detectors at the bottom of the Mediterranean Sea. The focus of ARCA is neutrino astronomy, while ORCA is optimised for neutrino oscillation studies. Both detectors are already operational in their intermediate states and collect valuable data, including the measurements of the muons produced by cosmic ray interactions in the atmosphere. This work explores the potential of machine learning models for the reconstruction of muon bundles, which are multi-muon events. For this, data collected with intermediate detector configurations of ARCA and ORCA was used in addition to simulated data from the envisaged final configurations of those detectors. Prediction of the total number of muons in a bundle as well as their total energy and even the energy of the primary cosmic ray is presented.

Research is a fundamental process driving the advancement of human civilization, yet it demands substantial time and effort from researchers. In recent years, the rapid development of artificial intelligence (AI) technologies has inspired researchers to explore how AI can accelerate and enhance research. To monitor relevant advancements, this paper presents a systematic review of the progress in this domain. Specifically, we organize the relevant studies into three main categories: hypothesis formulation, hypothesis validation, and manuscript publication. Hypothesis formulation involves knowledge synthesis and hypothesis generation. Hypothesis validation includes the verification of scientific claims, theorem proving, and experiment validation. Manuscript publication encompasses manuscript writing and the peer review process. Furthermore, we identify and discuss the current challenges faced in these areas, as well as potential future directions for research. Finally, we also offer a comprehensive overview of existing benchmarks and tools across various domains that support the integration of AI into the research process. We hope this paper serves as an introduction for beginners and fosters future research. Resources have been made publicly available at https://github.com/zkzhou126/AI-for-Research.

The development of social media platforms has revolutionized the speed and manner in which information is disseminated, leading to both beneficial and detrimental effects on society. While these platforms facilitate rapid communication, they also accelerate the spread of rumors and extremist speech, impacting public perception and behavior significantly. This issue is particularly pronounced during election periods, where the influence of social media on election outcomes has become a matter of global concern. With the unprecedented number of elections in 2024, against this backdrop, the election ecosystem has encountered unprecedented challenges. This study addresses the urgent need for effective rumor detection on social media by proposing a novel method that combines semantic analysis with graph neural networks. We have meticulously collected a dataset from PolitiFact and Twitter, focusing on politically relevant rumors. Our approach involves semantic analysis using a fine-tuned BERT model to vectorize text content and construct a directed graph where tweets and comments are nodes, and interactions are edges. The core of our method is a graph neural network, SAGEWithEdgeAttention, which extends the GraphSAGE model by incorporating first-order differences as edge attributes and applying an attention mechanism to enhance feature aggregation. This innovative approach allows for the fine-grained analysis of the complex social network structure, improving rumor detection accuracy. The study concludes that our method significantly outperforms traditional content analysis and time-based models, offering a theoretically sound and practically efficient solution.

Online Social Networks (OSN) are privileged observation channels for understanding the geospatial facets of many real-world phenomena [1]. Unfortunately, in most cases OSN content lacks explicit and structured geographic information, as in the case of Twitter, where only a minimal fraction (1% to 4%) of messages are natively geotagged [2]. This shortage of explicit geographic information drastically limits the exploitation of OSN data in geospatial Decision Support Systems (DSS) [3]. Conversely, the prompt availability of geotagged content would empower existing systems and would open up the possibility to develop new and better geospatial services and applications [4, 5]. As a practical example of this kind, several social media-based systems have been proposed in recent years for mapping and visualizing situational information in the aftermath of mass disasters - a task dubbed as crisis mapping - in an effort to augment emergency response [6, 7]. These systems, however, demand geotagged data to be placed on crisis maps, which in turn imposes to perform the geoparsing task on the majority of social media content. Explicit geographic information is not only needed in early warning [8, 9] and emergency response systems [10, 11, 12, 13, 14], but also in systems and applications for improving event promotion [15, 16], touristic planning [17, 18, 19], healthcare accessibility [20], news aggregation [21] Post-print of the article published in Decision Support Systems 136, 2020. Please refer to the published version: doi.org/10.1016/j.dss.2020.113346