Unslop

I ran the politicalcompass.org test on all major LLMs. They all landed in the lib-left quadrant. Grok as well, except when it didn't. There were two…

I scored the full text of over 12,000 arXiv papers from 2021 to 2026 across ten different fields and found that about a third of the papers currently…

training a 4b model with reinforcement learning to rewrite ai-sounding text so it reads human again - using ai detectors as the reward, and what happened when the model learned to hack them.

slople: one genuine human sentence a day. Rewrite it to sound as machine-written as you can. Unslop-MoE-v7 judges. Keep the facts, maximise the slop.

Fingers on the scale

You can find the winner and runners-up here. I am just starting to read the results. My impression so far is that this is not slop, but it's also not especially good. The same is true of much human fiction! Consistent with my earlier prediction that AI wi

Take the quiz: pick the AI paragraph from the human one. See how well you detect AI writing, and how our engine makes it harder to catch.

Fingers on the scale