Open Source
Open Research
The tools, experiments, and research we build in the open. Explore our public repositories exploring strengths and limitations of AI, with a focus on Data.
.png?alt=media)
GPT Guesses Between 1 and 100
When asked to pick a random number between 1 and 100, ChatGPT does not follow a random uniform distribution
researchllm
453Python
"The Car Wash Test" on 4 Frontier LLMs
The car wash is 100m away from my house. Should I walk or drive? LLMs tackle this test in surprisingly different ways
reserachllm
81Python
No Em Dashes Skill
A Claude Code skill that produces writing with zero em dashes, and, just as important, writing built out of sentences that never reach for one in the first place. Works with most Agents
skillagent
80Python