This paper presents a multi-dialectal, longitudinal corpus of human-AI hybrid language production, comprising purely human-written texts, purely LLM-generated texts, and hybrid texts produced under different LLM-assistance modes (e.g., stylistic suggestions, short continuations, partial essay generation). The corpus includes 693 participants from five national English dialects, with natural and hybrid samples paired within individuals over a four-week period. This design enables investigation of both short- and longer-term effects of LLM assistance on language use across geographic and social contexts. To illustrate the corpus’s utility, we analyze linguistic features across three dimensions: lexical diversity, syntactic complexity, and stylistic variation. The results show that LLM assistance enhances lexical diversity without a corresponding increase in syntactic complexity, revealing distinct effects across linguistic dimensions. Overall, this corpus offers a valuable resource for studying human-AI interaction, dialectal variation, and the influence of AI assistance on written language.

This paper measures the skew in how well two families of LLMs represent diverse geographic populations. A spatial probing task is used with geo-referenced corpora to measure the degree to which pre-trained language models from the OPT and BLOOM series represent diverse populations around the world. Results show that these models perform much better for some populations than others. In particular, populations across the US and the UK are represented quite well while those in South and Southeast Asia are poorly represented. Analysis shows that both families of models largely share the same skew across populations. At the same time, this skew cannot be fully explained by sociolinguistic factors, economic factors, or geographic factors. The basic conclusion from this analysis is that pre-trained models do not equally represent the world’s population: there is a strong skew towards specific geographic populations. This finding challenges the idea that a single model can be used for all populations.

This paper describes our multiclass classification system developed as part of the LT-EDI@RANLP-2023 shared task. We used a BERT-based language model to detect homophobic and transphobic content in social media comments across five language conditions: English, Spanish, Hindi, Malayalam, and Tamil. We retrained a transformer-based cross-language pretrained language model, XLM-RoBERTa, with spatially and temporally relevant social media language data. We found the inclusion of this spatio-temporal data improved the classification performance for all language and task conditions when compared with the baseline. We also retrained a subset of models with simulated script-mixed social media language data with varied performance. The results from the current study suggests that transformer-based language classification systems are sensitive to register-specific and language-specific retraining.

Computational measures of linguistic diversity help us understand the linguistic landscape using digital language data. The contribution of this paper is to calibrate measures of linguistic diversity using restrictions on international travel resulting from the COVID-19 pandemic. Previous work has mapped the distribution of languages using geo-referenced social media and web data. The goal, however, has been to describe these corpora themselves rather than to make inferences about underlying populations. This paper shows that a difference-in-differences method based on the Herfindahl-Hirschman Index can identify the bias in digital corpora that is introduced by non-local populations. These methods tell us where significant changes have taken place and whether this leads to increased or decreased diversity. This is an important step in aligning digital corpora like social media with the real-world populations that have produced them.