Hi all, I spent the past couple months building a factchecker to combat AI hallucination and sycophancy. The on-device sycophancy model is a Bert model and is still a prototype, and this Chrome extension is still not done yet, so I'd appreciate any feedback!