arXiv subjects

Damian J. Ruck

Publications and source records attributed to Damian J. Ruck.

3 recordsLinked to original sources

arXiv2026

Human--LLM Collaboration Is Transforming Complexity Metrics in Scientific Texts

While human language has long been studied as a complex system, Large Language Models (LLMs) are rapidly becoming contributors to its dynamics. Because LLMs are trained on human language use, their effects on the broader human-AI linguistic ecosystem are likely subtle at first. As their use becomes more widespread, however, LLMs may alter emergent properties of language, particularly as models increasingly train on mixed human-LLM textual data. Here, we draw on complexity science to look for subtle LLM effects in millions of arXiv abstracts from 2010 to 2025. The year 2023, when LLMs rapidly became widely used, serves as a landmark in a natural experiment. While we find a sharp increase in a composite LLM-associated style index after early 2023, we observe only subtle changes in the exponents of Zipf's law and Heaps' law. More compelling, however, are two subtle changes in complexity metrics that emerge from 2023 onward. First, turnover among top-ranked words increases sharply. Second, the positive relationship between the LLM-associated style index and three complexity metrics--vocabulary size and the exponents of Heaps' and Zipf's laws--becomes flatter after 2022. Together, these patterns are consistent with changes in the emergent properties of scientific text in a mixed human-AI linguistic ecosystem.

cs.CY↗

arXiv2026

Scalable Evaluation of the Realism of Synthetic Environmental Augmentations in Images

Evaluation of AI systems often requires synthetic test cases, particularly for rare or safety-critical conditions that are difficult to observe in operational data. Generative AI offers a promising approach for producing such data through controllable image editing, but its usefulness depends on whether the resulting images are sufficiently realistic to support meaningful evaluation. We present a scalable framework for assessing the realism of synthetic image-editing methods and apply it to the task of adding environmental conditions-fog, rain, snow, and nighttime-to car-mounted camera images. Using 40 clear-day images, we compare rule-based augmentation libraries with generative AI image-editing models. Realism is evaluated using two complementary automated metrics: a vision-language model (VLM) jury for perceptual realism assessment, and embedding-based distributional analysis to measure similarity to genuine adverse-condition imagery. Generative AI methods substantially outperform rule-based approaches, with the best generative method achieving approximately 3.6 times the acceptance rate of the best rule-based method. Performance varies across conditions: fog proves easiest to simulate, while nighttime transformations remain challenging. Notably, the VLM jury assigns imperfect acceptance even to real adverse-condition imagery, establishing practical ceilings against which synthetic methods can be judged. By this standard, leading generative methods match or exceed real-image performance for most conditions. These results suggest that modern generative image-editing models can enable scalable generation of realistic adverse-condition imagery for evaluation pipelines. Our framework therefore provides a practical approach for scalable realism evaluation, though validation against human studies remains an important direction for future work.

cs.CV↗

arXiv2022

Engagement Outweighs Exposure to Partisan and Unreliable News within Google Search

If popular online platforms systematically expose their users to partisan and unreliable news, they could potentially contribute to societal issues like rising political polarization. This concern is central to the echo chamber and filter bubble debates, which critique the roles that user choice and algorithmic curation play in guiding users to different online information sources. These roles can be measured in terms of exposure, the URLs seen while using an online platform, and engagement, the URLs selected while on that platform or browsing the web more generally. However, due to the challenges of obtaining ecologically valid exposure data--what real users saw during their regular platform use--studies in this vein often only examine engagement data, or estimate exposure via simulated behavior or inference. Despite their centrality to the contemporary information ecosystem, few such studies have focused on web search, and even fewer have examined both exposure and engagement on any platform. To address these gaps, we conducted a two-wave study pairing surveys with ecologically valid measures of exposure and engagement on Google Search during the 2018 and 2020 US elections. We found that participants' partisan identification had a small and inconsistent relationship with the amount of partisan and unreliable news they were exposed to on Google Search, a more consistent relationship with the search results they chose to follow, and the most consistent relationship with their overall engagement. That is, compared to the news sources our participants were exposed to on Google Search, we found more identity-congruent and unreliable news sources in their engagement choices, both within Google Search and overall. These results suggest that exposure and engagement with partisan or unreliable news on Google Search are not primarily driven by algorithmic curation, but by users' own choices.

cs.SI↗