I Trained a Markdown File to Boost GPT-5.5 by 23 Points — It Shouldn’t Work (opens in new tab)

Last Updated on June 18, 2026 by Editorial Team Author(s): Chew Loong Nian – AI ENGINEER Originally published on Towards AI. I Trained a Markdown File to Boost GPT-5.5 by 23 Points — It Shouldn’t Work I did not fine-tune anything. I did not touch a single weight. I ran a training loop whose only output was a 1,400-token Markdown file, dropped that file into the context window of a frozen GPT-5.5, and watched its six-benchmark average jump from 58.8 to 82.3. That is +23.5 points from a text fi...