I'm Jared! I'm an engineer and social impact entrepreneur. I love drinking coffee, building robots, and hunting for adventure.
Below is a collection of things I've created - some at work, some on the side. If you'd like to talk shop, or just hear a little more about one of my projects, get in touch.
ð Social Impact ðą Sustainability ð ïļ Engineering ð Adventure
Safety evals may prove critical for catching misaligned behavior before a model is released. However, LLMs increasingly display awareness that they are being evaluated. This eval-awareness has the potential to quickly render evals useless.
My current research: What Makes an Eval Look Fake? A Controlled Ablation Study of Realism Judge Sensitivity
View project âBuilding an agent to triage incoming support requests and troubleshoot issues in the IoT device net.
Visit Smartex.ai âBringing the iOS app to Android.
Visit presh.app âAre you working in AI safety or alternative protiens?
Let's chat â