Blog
6 posts
The Model Is a Database. You Just Couldn't Query It Until Now.
LARQL decompiles transformer weights into a queryable graph. SQL over FFN knowledge. Insert facts without retraining. Inspect what the model actually knows.
Read article →Body Snatching: Growing a New Model Inside Someone Else's Skeleton
Full fine-tuning replaces a model's identity but costs $10,000. Progressive LoRA Merging does the same thing on one GPU for under $500.
Read article →ASRL: Why Supervised and Reinforcement Learning Should Take Turns
Everyone trains SFT first, then RL. I alternated them inside each epoch. The model converged in 3 epochs instead of 12.
Read article →OpenClaw's Memory Problem: Why Your Agent Can't Keep Promises
AI agents promise to remember. They can't. Here's why context compaction and stateless sessions break every memory system, and how a compaction agent plus QMD fixes it.
Read article →I Ranked Next to Wikipedia in ChatGPT. Here's What Actually Happened.
A solo experiment to prove SEO is about novelty, not rules. How Toxigon ended up at 4.1% of ChatGPT global citations alongside Wikipedia, Reddit, and Forbes.
Read article →Consciousness Is Not a Feature You Can Deploy
People keep conflating intelligence with consciousness. They are not the same thing, and no amount of compute will close that gap.
Read article →