Sitemap - 2025 - The Counterfactual
We should be cautious about LLM-generated benchmarks
How I use (and don't use) ChatGPT
Why isn't language more iconic?
When words sound (or look) like what they mean
Informed consent is central to research ethics
Identifying signatures of LLM-generated text
LLM-ology and the "moving target" problem
An overlooked problem with LLM benchmarks