Mathos AI (formerly MathGPTPro). Best AI Math Tutor towards Personalized Learning.Snap a photo math and ask & learn. Instantly sync across all platforms, available on the web app, iOS, and Andriod.𝟭𝟱% accuracy improvement over GPT4o.𝗢𝘃𝗲𝗿 𝟭𝗠 users through organic growth.𝗛𝗜𝗥𝗜𝗡𝗚: MKT interns, LLM interns/full-time, 𝗦𝗿. 𝗙𝗿𝗼𝗻𝘁𝗲𝗻𝗱 𝗥𝗲𝗮𝗰𝘁, Fullstack (full-time) (remote or hybrid: we're flexible as long as we share the same passion)We shift 𝗳𝗮𝘀𝘁. We 𝗽𝘂𝗯𝗹𝗶𝘀𝗵 𝗿𝗲𝘀𝗲𝗮𝗿𝗰𝗵 paper on Tier 1 AI LLM conferences. We develop LLM models and build user-friendly applications. So far, our website reached 𝗺𝗶𝗹𝗹𝗶𝗼𝗻𝘀 of monthly views with 𝗣𝗟𝗚. So if you have expertise in 𝗴𝗿𝗼𝘄𝘁𝗵, we believe you can bring our traction to the next level.

K-Scale Labs (YC W24, AI Grant 3)Hey Revia (YC S24)Hamming AI (YC S24, AI Grant 4)Story Engine (a16z Speedrun)AG2 (Github Stars 20k Agent OS)

Recommendation

Deep Reinforcement Learning, 19 Fall: https://cmudeeprl.github.io/703website/

Exploration-exploitation decoposion for curious Model-Agnostic Meta-Learning (MAML), with posterior sampling accounting for task uncertainty inference.