Recent advancements in Large Language Models (LLMs) have empowered LLM agents to autonomously collect world information, over which to conduct reasoning to solve complex problems. Given this capability, increasing interests have been put into employing LLM agents for predicting international events, which can influence decision-making and shape policy development on an international scale. Despite such a growing interest, there is a lack of a rigorous benchmark of LLM agents' forecasting capability and reliability. To address this gap, we introduce MIRAI, a novel benchmark designed to systematically evaluate LLM agents as temporal forecasters in the context of international events. Our benchmark features an agentic environment with tools for accessing an extensive database of historical, structured events and textual news articles. We refine the GDELT event database with careful cleaning and parsing to curate a series of relational prediction tasks with varying forecasting horizons, assessing LLM agents' abilities from short-term to long-term forecasting. We further implement APIs to enable LLM agents to utilize different tools via a code-based interface. In summary, MIRAI comprehensively evaluates the agents' capabilities in three dimensions: 1) autonomously source and integrate critical information from large global databases; 2) write codes using domain-specific APIs and libraries for tool-use; and 3) jointly reason over historical knowledge from diverse formats and time to accurately predict future events. Through comprehensive benchmarking, we aim to establish a reliable framework for assessing the capabilities of LLM agents in forecasting international events, thereby contributing to the development of more accurate and trustworthy models for international relation analysis.

It has become routine to report research results where Large Language Models (LLMs) outperform average humans in a wide range of language-related tasks, and creative text writing is no exception. It seems natural, then, to raise the bid: Are LLMs ready to compete in creative writing skills with a top (rather than average) novelist? To provide an initial answer for this question, we have carried out a contest between Patricio Pron (an awarded novelist, considered one of the best of his generation) and GPT-4 (one of the top performing LLMs), in the spirit of AI-human duels such as DeepBlue vs Kasparov and AlphaGo vs Lee Sidol. We asked Pron and GPT-4 to provide thirty titles each, and then to write short stories for both their titles and their opponent's. Then, we prepared an evaluation rubric inspired by Boden's definition of creativity, and we collected 5,400 manual assessments provided by literature critics and scholars. The results of our experimentation indicate that LLMs are still far from challenging a top human creative writer, and that reaching such level of autonomous creative writing skills probably cannot be reached simply with larger language models.

This paper introduces a collaborative, human-centered taxonomy of AI, algorithmic and automation harms. We argue that existing taxonomies, while valuable, can be narrow, unclear, typically cater to practitioners and government, and often overlook the needs of the wider public. Drawing on existing taxonomies and a large repository of documented incidents, we propose a taxonomy that is clear and understandable to a broad set of audiences, as well as being flexible, extensible, and interoperable. Through iterative refinement with topic experts and crowdsourced annotation testing, we propose a taxonomy that can serve as a powerful tool for civil society organisations, educators, policymakers, product teams and the general public. By fostering a greater understanding of the real-world harms of AI and related technologies, we aim to increase understanding, empower NGOs and individuals to identify and report violations, inform policy discussions, and encourage responsible technology development and deployment.

In this paper, we apply a method to quantify biases associated with named entities from various countries. We create counterfactual examples with small perturbations on target-domain data instead of relying on templates or specific datasets for bias detection. On widely used classifiers for subjectivity analysis, including sentiment, emotion, hate speech, and offensive text using Twitter data, our results demonstrate positive biases related to the language spoken in a country across all classifiers studied. Notably, the presence of certain country names in a sentence can strongly influence predictions, up to a 23\% change in hate speech detection and up to a 60\% change in the prediction of negative emotions such as anger. We hypothesize that these biases stem from the training data of pre-trained language models (PLMs) and find correlations between affect predictions and PLMs likelihood in English and unknown languages like Basque and Maori, revealing distinct patterns with exacerbate correlations. Further, we followed these correlations in-between counterfactual examples from a same sentence to remove the syntactical component, uncovering interesting results suggesting the impact of the pre-training data was more important for English-speaking-country names. Our anonymized code is [https://anonymous.4open.science/r/biases_ppl-576B/README.md](available here).

AI regulations are expected to prohibit machine learning models from using sensitive attributes during training. However, the latest Natural Language Processing (NLP) classifiers, which rely on deep learning, operate as black-box systems, complicating the detection and remediation of such misuse. Traditional bias mitigation methods in NLP aim for comparable performance across different groups based on attributes like gender or race but fail to address the underlying issue of reliance on protected attributes. To partly fix that, we introduce NLPGuard, a framework for mitigating the reliance on protected attributes in NLP classifiers. NLPGuard takes an unlabeled dataset, an existing NLP classifier, and its training data as input, producing a modified training dataset that significantly reduces dependence on protected attributes without compromising accuracy. NLPGuard is applied to three classification tasks: identifying toxic language, sentiment analysis, and occupation classification. Our evaluation shows that current NLP classifiers heavily depend on protected attributes, with up to $23\%$ of the most predictive words associated with these attributes. However, NLPGuard effectively reduces this reliance by up to $79\%$, while slightly improving accuracy.

Advancements in Large Language Models (LLMs) have significantly enhanced instruction-following capabilities. However, most Instruction Fine-Tuning (IFT) datasets are predominantly in English, limiting model performance in other languages. Traditional methods for creating multilingual IFT datasets such as translating existing English IFT datasets or converting existing NLP datasets into IFT datasets by templating, struggle to capture linguistic nuances and ensure prompt (instruction) diversity. To address this issue, we propose a novel method for collecting multilingual IFT datasets that preserves linguistic naturalness and ensures prompt diversity. This approach leverages English-focused LLMs, monolingual corpora, and a scoring function to create high-quality, diversified IFT datasets in multiple languages. Experiments demonstrate that LLMs finetuned using these IFT datasets show notable improvements in both generative and discriminative tasks, indicating enhanced language comprehension by LLMs in non-English contexts. Specifically, on the multilingual summarization task, LLMs using our IFT dataset achieved 17.57% and 15.23% improvements over LLMs fine-tuned with translation-based and template-based datasets, respectively.

Generalization and robustness to input variation are core desiderata of machine learning research. Language varies along several axes, most importantly, language instance (e.g. French) and domain (e.g. news). While adapting NLP models to new languages within a single domain, or to new domains within a single language, is widely studied, research in joint adaptation is hampered by the lack of evaluation datasets. This prevents the transfer of NLP systems from well-resourced languages and domains to non-dominant language-domain combinations. To address this gap, we introduce M2QA, a multi-domain multilingual question answering benchmark. M2QA includes 13,500 SQuAD 2.0-style question-answer instances in German, Turkish, and Chinese for the domains of product reviews, news, and creative writing. We use M2QA to explore cross-lingual cross-domain performance of fine-tuned models and state-of-the-art LLMs and investigate modular approaches to domain and language adaptation. We witness 1) considerable performance variations across domain-language combinations within model classes and 2) considerable performance drops between source and target language-domain combinations across all model sizes. We demonstrate that M2QA is far from solved, and new methods to effectively transfer both linguistic and domain-specific information are necessary. We make M2QA publicly available at https://github.com/UKPLab/m2qa.

Opinion Dynamics is an interdisciplinary area of research. Psychology and Sociology have proposed models of how individuals form opinions and how social interactions influence this process. Socio-Physicists have interpreted patterns in opinion formation as arising from non-linearity in the underlying process, shaping the models. Agent-based modeling has offered a platform to study the Opinion Dynamics of large groups. This paper recasts recent models in opinion formation into a proper dynamical system, injecting the idea of clock time into evolving opinions. The time interval between successive receipts of new information (frequency of information receipts) becomes a factor to study. Social media has shrunk time intervals between information receipts, increasing their frequency. The recast models show that shorter intervals and larger networks increase an individual's propensity for polarization, defined as an inability to hold a neutral opinion. A Polarization number based on sociological parameters is proposed, with critical values beyond which individuals are prone to polarization, depending on psychological parameters. Reduced time intervals and larger interacting groups can push the Polarization number to critical values, contributing to polarization. The Extent of Polarization is defined as the width of the region around neutral within which an individual cannot hold an opinion. Results are reported for model parameters found in the literature. The findings offer an opportunity to adjust model parameters to align with empirical evidence, aiding the study of Opinion Dynamics in large social networks using Agent-Based Modeling.

The rapid evolution of deep learning architectures has significantly advanced the field of computer vision, particularly in tasks that require the analysis of complex spatial data. Convolutional Neural Networks (CNNs), initially proposed by LeCun et al. [1], have become a cornerstone in this domain due to their ability to efficiently process highdimensional data arrays such as images. These networks typically employ linear transformations followed by activation functions in their convolutional layers to discern spatial relationships, thereby reducing the number of parameters needed to capture intricate patterns in visual data. Since 2012, following the success of AlexNet [2] in the ImageNet classification challenge, CNNs have dominated the field of computer vision until the emergence of Vision Transformers [3]. Innovations such as Residual Networks [4] and Densely Connected networks [5], along with numerous subsequent works, have significantly advanced the achievable quality of models based on convolutional layers, enabling the effective training of very large and deep networks. In segmentation tasks, especially within the biomedical domain, CNNs have also become foundational with the advent of the U-Net [6] architecture, which has subsequently inspired a whole family of U-Net-like architectures for segmentation tasks. Recent developments in deep learning have seen the integration of sophisticated mathematical theories into neural network architectures, enhancing their capability to handle complex data structures. One such innovation is the Kolmogorov-Arnold Network (KAN) [7], which leverages the Kolmogorov-Arnold theorem to incorporate splines into its architecture, offering a compelling alternative to traditional Multi-Layer Perceptrons (MLPs).

This paper presents the POLygraph dataset, a unique resource for fake news detection in Polish. The dataset, created by an interdisciplinary team, is composed of two parts: the "fake-or-not" dataset with 11,360 pairs of news articles (identified by their URLs) and corresponding labels, and the "fake-they-say" dataset with 5,082 news articles (identified by their URLs) and tweets commenting on them. Unlike existing datasets, POLygraph encompasses a variety of approaches from source literature, providing a comprehensive resource for fake news detection. The data was collected through manual annotation by expert and non-expert annotators. The project also developed a software tool that uses advanced machine learning techniques to analyze the data and determine content authenticity. The tool and dataset are expected to benefit various entities, from public sector institutions to publishers and fact-checking organizations. Further dataset exploration will foster fake news detection and potentially stimulate the implementation of similar models in other languages. The paper focuses on the creation and composition of the dataset, so it does not include a detailed evaluation of the software tool for content authenticity analysis, which is planned at a later stage of the project.