arXiv subjects

David Garcia

Publications and source records attributed to David Garcia.

At least 37 records · Page 2Linked to original sources

arXiv2025

AI agents can coordinate beyond human scale

Large language models (LLMs) are increasingly deployed in collaborative tasks involving multiple agents, forming an "AI agent society: where agents interact and influence one another. Whether such groups can spontaneously coordinate on arbitrary decisions without external influence - a hallmark of self-organized regulation in human societies - remains an open question. Here we investigate the stability of groups formed by AI agents by applying methods from complexity science and principles from behavioral sciences. We find that LLMs can spontaneously form cohesive groups, and that their opinion dynamics is governed by a majority force coefficient, which determines whether coordination is achievable. This majority force diminishes as group size increases, leading to a critical group size beyond which coordination becomes practically unattainable and stability is lost. Notably, this critical group size grows exponentially with the language capabilities of the models, and for the most advanced LLMs, it exceeds the typical size of informal human groups. Our findings highlight intrinsic limitations in the self-organization of AI agent societies and have implications for the design of collaborative AI systems where coordination is desired or could represent a treat.

physics.soc-ph↗

arXiv2025

R.U.Psycho? Robust Unified Psychometric Testing of Language Models

Generative language models are increasingly being subjected to psychometric questionnaires intended for human testing, in efforts to establish their traits, as benchmarks for alignment, or to simulate participants in social science experiments. While this growing body of work sheds light on the likeness of model responses to those of humans, concerns are warranted regarding the rigour and reproducibility with which these experiments may be conducted. Instabilities in model outputs, sensitivity to prompt design, parameter settings, and a large number of available model versions increase documentation requirements. Consequently, generalization of findings is often complex and reproducibility is far from guaranteed. In this paper, we present R.U.Psycho, a framework for designing and running robust and reproducible psychometric experiments on generative language models that requires limited coding expertise. We demonstrate the capability of our framework on a variety of psychometric questionnaires, which lend support to prior findings in the literature. R.U.Psycho is available as a Python package at https://github.com/julianschelb/rupsycho.

cs.CL↗

arXiv2025

Extracting Affect Aggregates from Longitudinal Social Media Data with Temporal Adapters for Large Language Models

This paper proposes temporally aligned Large Language Models (LLMs) as a tool for longitudinal analysis of social media data. We fine-tune Temporal Adapters for Llama 3 8B on full timelines from a panel of British Twitter users, and extract longitudinal aggregates of emotions and attitudes with established questionnaires. We focus our analysis on the beginning of the COVID-19 pandemic that had a strong impact on public opinion and collective emotions. We validate our estimates against representative British survey data and find strong positive, significant correlations for several collective emotions. The obtained estimates are robust across multiple training seeds and prompt formulations, and in line with collective emotions extracted using a traditional classification model trained on labeled data. We demonstrate the flexibility of our method on questions of public opinion for which no pre-trained classifier is available. Our work extends the analysis of affect in LLMs to a longitudinal setting through Temporal Adapters. It enables flexible, new approaches towards the longitudinal analysis of social media data.

cs.CY↗

arXiv2024

Computational analysis of US Congressional speeches reveals a shift from evidence to intuition

Pursuit of honest and truthful decision-making is crucial for governance and accountability in democracies. However, people sometimes take different perspectives of what it means to be honest and how to pursue truthfulness. Here we explore a continuum of perspectives from evidence-based reasoning, rooted in ascertainable facts and data, at one end, to intuitive decisions that are driven by feelings and subjective interpretations, at the other. We analyze the linguistic traces of those contrasting perspectives in Congressional speeches from 1879 to 2022. We find that evidence-based language has continued to decline since the mid-1970s, together with a decline in legislative productivity. The decline was accompanied by increasing partisan polarization in Congress and rising income inequality in society. Results highlight the importance of evidence-based language in political decision-making.

econ.GN↗

arXiv2024

Unpacking polarization: Antagonism and Alignment in Signed Networks of Online Interaction

Political conflict is an essential element of democratic systems, but can also threaten their existence if it becomes too intense. This happens particularly when most political issues become aligned along the same major fault line, splitting society into two antagonistic camps. In the 20th century, major fault lines were formed by structural conflicts, like owners vs workers, center vs periphery, etc. But these classical cleavages have since lost their explanatory power. Instead of theorizing new cleavages, we present the FAULTANA (FAULT-line Alignment Network Analysis) pipeline, a computational method to uncover major fault lines in data of signed online interactions. Our method makes it possible to quantify the degree of antagonism prevalent in different online debates, as well as how aligned each debate is to the major fault line. This makes it possible to identify the wedge issues driving polarization, characterized by both intense antagonism and alignment. We apply our approach to large-scale data sets of Birdwatch, a US-based Twitter fact-checking community and the discussion forums of DerStandard, an Austrian online newspaper. We find that both online communities are divided into two large groups and that their separation follows political identities and topics. In addition, for DerStandard, we pinpoint issues that reinforce societal fault lines and thus drive polarization. We also identify issues that trigger online conflict without strictly aligning with those dividing lines (e.g. COVID-19). Our methods allow us to construct a time-resolved picture of affective polarization that shows the separate contributions of cohesiveness and divisiveness to the dynamics of alignment during contentious elections and events.

cs.SI↗

arXiv2023

Emergence of Scale-Free Networks in Social Interactions among Large Language Models

Scale-free networks are one of the most famous examples of emergent behavior and are ubiquitous in social systems, especially online social media in which users can follow each other. By analyzing the interactions of multiple generative agents using GPT3.5-turbo as a language model, we demonstrate their ability to not only mimic individual human linguistic behavior but also exhibit collective phenomena intrinsic to human societies, in particular the emergence of scale-free networks. We discovered that this process is disrupted by a skewed token prior distribution of GPT3.5-turbo, which can lead to networks with extreme centralization as a kind of alignment. We show how renaming agents removes these token priors and allows the model to generate a range of networks from random networks to more realistic scale-free networks.

physics.soc-ph↗

arXiv2023

From alternative conceptions of honesty to alternative facts in communications by U.S. politicians

The spread of online misinformation on social media is increasingly perceived as a problem for societal cohesion and democracy. The role of political leaders in this process has attracted less research attention, even though politicians who "speak their mind" are perceived by segments of the public as authentic and honest even if their statements are unsupported by evidence. Analyzing communications by members of the U.S. Congress on Twitter between 2011 and 2022, we show that politicians' conception of honesty has undergone a distinct shift, with authentic belief-speaking that may be decoupled from evidence becoming more prominent and more differentiated from explicitly evidence-based truth seeking. We show that for Republicans - but not Democrats - an increase of belief-speaking of 10% is associated with a decrease of 12.8 points of quality (NewsGuard scoring system) in the sources shared in a tweet. Conversely, an increase in truth-seeking language is associated with an increase in quality of sources for both parties. The results support the hypothesis that the current dissemination of misinformation in political discourse is in part driven by an alternative understanding of truth and honesty that emphasizes invocation of subjective belief at the expense of reliance on evidence.

cs.CY↗

arXiv2023

LEIA: Linguistic Embeddings for the Identification of Affect

The wealth of text data generated by social media has enabled new kinds of analysis of emotions with language models. These models are often trained on small and costly datasets of text annotations produced by readers who guess the emotions expressed by others in social media posts. This affects the quality of emotion identification methods due to training data size limitations and noise in the production of labels used in model development. We present LEIA, a model for emotion identification in text that has been trained on a dataset of more than 6 million posts with self-annotated emotion labels for happiness, affection, sadness, anger, and fear. LEIA is based on a word masking method that enhances the learning of emotion words during model pre-training. LEIA achieves macro-F1 values of approximately 73 on three in-domain test datasets, outperforming other supervised and unsupervised methods in a strong benchmark that shows that LEIA generalizes across posts, users, and time periods. We further perform an out-of-domain evaluation on five different datasets of social media and other sources, showing LEIA's robust performance across media, data collection methods, and annotation schemes. Our results show that LEIA generalizes its classification of anger, happiness, and sadness beyond the domain it was trained on. LEIA can be applied in future research to provide better identification of emotions in text from the perspective of the writer. The models produced for this article are publicly available at https://huggingface.co/LEIA

cs.CL↗

arXiv2023

Neural Lumped Parameter Differential Equations with Application in Friction-Stir Processing

Lumped parameter methods aim to simplify the evolution of spatially-extended or continuous physical systems to that of a "lumped" element representative of the physical scales of the modeled system. For systems where the definition of a lumped element or its associated physics may be unknown, modeling tasks may be restricted to full-fidelity simulations of the physics of a system. In this work, we consider data-driven modeling tasks with limited point-wise measurements of otherwise continuous systems. We build upon the notion of the Universal Differential Equation (UDE) to construct data-driven models for reducing dynamics to that of a lumped parameter and inferring its properties. The flexibility of UDEs allow for composing various known physical priors suitable for application-specific modeling tasks, including lumped parameter methods. The motivating example for this work is the plunge and dwell stages for friction-stir welding; specifically, (i) mapping power input into the tool to a point-measurement of temperature and (ii) using this learned mapping for process control.

cs.LG↗

arXiv2023

This Sample seems to be good enough! Assessing Coverage and Temporal Reliability of Twitter's Academic API

Because of its willingness to share data with academia and industry, Twitter has been the primary social media platform for scientific research as well as for consulting businesses and governments in the last decade. In recent years, a series of publications have studied and criticized Twitter's APIs and Twitter has partially adapted its existing data streams. The newest Twitter API for Academic Research allows to "access Twitter's real-time and historical public data with additional features and functionality that support collecting more precise, complete, and unbiased datasets." The main new feature of this API is the possibility of accessing the full archive of all historic Tweets. In this article, we will take a closer look at the Academic API and will try to answer two questions. First, are the datasets collected with the Academic API complete? Secondly, since Twitter's Academic API delivers historic Tweets as represented on Twitter at the time of data collection, we need to understand how much data is lost over time due to Tweet and account removal from the platform. Our work shows evidence that Twitter's Academic API can indeed create (almost) complete samples of Twitter data based on a wide variety of search terms. We also provide evidence that Twitter's data endpoint v2 delivers better samples than the previously used endpoint v1.1. Furthermore, collecting Tweets with the Academic API at the time of studying a phenomenon rather than creating local archives of stored Tweets, allows for a straightforward way of following Twitter's developer agreement. Finally, we will also discuss technical artifacts and implications of the Academic API. We hope that our work can add another layer of understanding of Twitter data collections leading to more reliable studies of human behavior via social media data.

cs.SI↗

arXiv2023

Just Another Day on Twitter: A Complete 24 Hours of Twitter Data

At the end of October 2022, Elon Musk concluded his acquisition of Twitter. In the weeks and months before that, several questions were publicly discussed that were not only of interest to the platform's future buyers, but also of high relevance to the Computational Social Science research community. For example, how many active users does the platform have? What percentage of accounts on the site are bots? And, what are the dominating topics and sub-topical spheres on the platform? In a globally coordinated effort of 80 scholars to shed light on these questions, and to offer a dataset that will equip other researchers to do the same, we have collected all 375 million tweets published within a 24-hour time period starting on September 21, 2022. To the best of our knowledge, this is the first complete 24-hour Twitter dataset that is available for the research community. With it, the present work aims to accomplish two goals. First, we seek to answer the aforementioned questions and provide descriptive metrics about Twitter that can serve as references for other researchers. Second, we create a baseline dataset for future research that can be used to study the potential impact of the platform's ownership change.

cs.SI↗

arXiv2022

Social media sharing by political elites: An asymmetric American exceptionalism

Increased sharing of untrustworthy information on social media platforms is one of the main challenges of our modern information society. Because information disseminated by political elites is known to shape citizen and media discourse, it is particularly important to examine the quality of information shared by politicians. Here we show that from 2016 onward, members of the Republican party in the U.S. Congress have been increasingly sharing links to untrustworthy sources. The proportion of untrustworthy information posted by Republicans versus Democrats is diverging at an accelerating rate, and this divergence has worsened since president Biden was elected. This divergence between parties seems to be unique to the U.S. as it cannot be observed in other western democracies such as Germany and the United Kingdom, where left-right disparities are smaller and have remained largely constant.

cs.CY↗

arXiv2022

Detecting potentially harmful and protective suicide-related content on twitter: A machine learning approach

Research shows that exposure to suicide-related news media content is associated with suicide rates, with some content characteristics likely having harmful and others potentially protective effects. Although good evidence exists for a few selected characteristics, systematic large scale investigations are missing in general, and in particular for social media data. We apply machine learning methods to classify large quantities of Twitter data according to a novel annotation scheme that distinguishes 12 categories of suicide-related tweets. We then trained a benchmark of machine learning models including a majority classifier, an approach based on word frequency (TF-IDF with a linear SVM) and two state-of-the-art deep learning models (BERT, XLNet). The two deep learning models achieved the best performance in two classification tasks: In the first task, we classified six main content categories, including personal stories about either suicidal ideation and attempts or coping, calls for action intending to spread either problem awareness or prevention-related information, reporting of suicide cases, and other tweets irrelevant to these categories. The deep learning models reached accuracy scores above 73% on average across the six categories, and F1-scores in between 0.70 and 0.85 for all but the suicidal ideation and attempts category (0.51-0.55). In the second task, separating tweets referring to actual suicide from off-topic tweets, they correctly labeled around 88% of tweets, with BERT achieving F1-scores of 0.93 and 0.74 for the two categories, respectively. These classification performances are comparable to the state-of-the-art on similar tasks. By making data labeling more efficient, this work has enabled large-scale investigations on harmful and protective associations of social media content with suicide rates and help-seeking behavior.

cs.CL↗

arXiv2022

LEXpander: applying colexification networks to automated lexicon expansion

Recent approaches to text analysis from social media and other corpora rely on word lists to detect topics, measure meaning, or to select relevant documents. These lists are often generated by applying computational lexicon expansion methods to small, manually-curated sets of root words. Despite the wide use of this approach, we still lack an exhaustive comparative analysis of the performance of lexicon expansion methods and how they can be improved with additional linguistic data. In this work, we present LEXpander, a method for lexicon expansion that leverages novel data on colexification, i.e. semantic networks connecting words based on shared concepts and translations to other languages. We evaluate LEXpander in a benchmark including widely used methods for lexicon expansion based on various word embedding models and synonym networks. We find that LEXpander outperforms existing approaches in terms of both precision and the trade-off between precision and recall of generated word lists in a variety of tests. Our benchmark includes several linguistic categories and sentiment variables in English and German. We also show that the expanded word lists constitute a high-performing text analysis method in application cases to various corpora. This way, LEXpander poses a systematic automated solution to expand short lists of words into exhaustive and accurate word lists that can closely approximate word lists generated by experts in psychology and linguistics.

cs.CL↗

arXiv2022

Validating daily social media macroscopes of emotions

To study emotions at the macroscopic level, affective scientists have made extensive use of sentiment analysis on social media text. However, this approach can suffer from a series of methodological issues with respect to sampling biases and measurement error. To date, it has not been validated if social media sentiment can measure the day to day temporal dynamics of emotions aggregated at the macro level of a whole online community. We ran a large-scale survey at an online newspaper to gather daily self-reports of affective states from its users and compare these with aggregated results of sentiment analysis of user discussions on the same online platform. Additionally, we preregistered a replication of our study using Twitter text as a macroscope of emotions for the same community. For both platforms, we find strong correlations between text analysis results and levels of self-reported emotions, as well as between inter-day changes of both measurements. We further show that a combination of supervised and unsupervised text analysis methods is the most accurate approach to measure emotion aggregates. We illustrate the application of such social media macroscopes when studying the association between the number of new COVID-19 cases and emotions, showing that the strength of associations is comparable when using survey data as when using social media data. Our findings indicate that macro level dynamics of affective states of users of an online platform can be tracked with social media text, complementing surveys when self-reported data is not available or difficult to gather.

cs.SI↗

arXiv2022

Assessment of the effectiveness of Omicron transmission mitigation strategies for European universities using an agent-based network model

Returning universities to full on-campus operations while the COVID-19 pandemic is ongoing has been a controversial discussion in many countries. The risk of large outbreaks in dense course settings is contrasted by the benefits of in-person teaching. Transmission risk depends on a range of parameters, such as vaccination coverage and efficacy, number of contacts and adoption of non-pharmaceutical intervention measures (NPIs). Due to the generalised academic freedom in Europe, many universities are asked to autonomously decide on and implement intervention measures and regulate on-campus operations. In the context of rapidly changing vaccination coverage and parameters of the virus, universities often lack sufficient scientific insight to base these decisions on. To address this problem, we analyse a calibrated, data-driven agent-based simulation of transmission dynamics of 10755 students and 974 faculty members in a medium-sized European university. We use a co-location network reconstructed from student enrollment data and calibrate transmission risk based on outbreak size distributions in education institutions. We focus on actionable interventions that are part of the already existing decision-making process of universities to provide guidance for concrete policy decisions. Here we show that, with the Omicron variant of the SARS-CoV-2 virus, even a reduction to 25% occupancy and universal mask mandates are not enough to prevent large outbreaks given the vaccination coverage of about 80% recently reported for students in Austria. Our results show that controlling the spread of the virus with available vaccines in combination with NPIs is not feasible in the university setting if presence of students and faculty on campus is required.

physics.soc-ph↗

arXiv2021

Social media emotion macroscopes reflect emotional experiences in society at large

Social media generate data on human behaviour at large scales and over long periods of time, posing a complementary approach to traditional methods in the social sciences. Millions of texts from social media can be processed with computational methods to study emotions over time and across regions. However, recent research has shown weak correlations between social media emotions and affect questionnaires at the individual level and between static regional aggregates of social media emotion and subjective well-being at the population level, questioning the validity of social media data to study emotions. Yet, to date, no research has tested the validity of social media emotion macroscopes to track the temporal evolution of emotions at the level of a whole society. Here we present a pre-registered prediction study that shows how gender-rescaled time series of Twitter emotional expression at the national level substantially correlate with aggregates of self-reported emotions in a weekly representative survey in the United Kingdom. A follow-up exploratory analysis shows a high prevalence of third-person references in emotionally-charged tweets, indicating that social media data provide a way of social sensing the emotions of others rather than just the emotional experiences of users. These results show that, despite the issues that social media have in terms of representativeness and algorithmic confounding, the combination of advanced text analysis methods with user demographic information in social media emotion macroscopes can provide measures that are informative of the general population beyond social media users.

cs.SI↗

arXiv2020

Fragile, yet resilient: Adaptive decline in a collaboration network of firms

The dynamics of collaboration networks of firms follow a life-cycle of growth and decline. That does not imply they also become less resilient. Instead, declining collaboration networks may still have the ability to mitigate shocks from firms leaving, and to recover from these losses by adapting to new partners. To demonstrate this, we analyze 21.500 R\&D collaborations of 14.500 firms in six different industrial sectors over 25 years. We calculate time-dependent probabilities of firms leaving the network and simulate drop-out cascades, to determine the expected dynamics of decline. We then show that deviations from these expectations result from the adaptivity of the network, which mitigates the decline. These deviations can be used as a measure of network resilience.

physics.soc-ph↗