Research Motivation
This project implements research on AI-associated lexical overuse. It is motivated by three simple observations:
- AI systems exhibit a distinct language style;
- millions of people interact with AI-generated language every day; and
- sustained exposure to this language may influence how humans write and speak.
A central goal of this line of work is to develop a fully automated procedure for identifying words that are systematically overused by large language models. Automation makes it possible to scale the analysis across many languages, domains, and model architectures without manual curation. The Explorer makes this procedure accessible and interactive.
Methodological Focus
The underlying methodology compares AI-generated continuations with matched human baselines and tracks lexical prevalence across documents. This makes it possible to identify words that appear disproportionately often in AI output.
The broader aim is not only to identify which words are overused, but also to understand the mechanisms behind these patterns and how they vary across settings.
The “Why”
This research goes beyond identifying which words are overused. It asks why large language models develop language styles that diverge so strongly from human baselines.
A substantial part of this divergence appears to be linked to Reinforcement Learning from Human Feedback (RLHF). However, many aspects remain underexplored, including the role of annotator demographics, the influence of task framing, and the effects of different optimisation objectives on lexical behaviour. Understanding these mechanisms is important because millions of people are exposed to AI output and because linguistic alignment is only one part of the broader alignment problem.
Project Link
Repository: Anonymous project page