Is It EVEN Possible To Reverse Engineer AI’s Training Data?

Deploy on Sevalla now and get a free $50 credit! https://sevalla.com/?utm_source=ByCloud's&utm_medium=Referral&utm_campaign=youtube In this video, we dive into how much of the private training data researchers can infer or approximate. My Newsletter https://mail.bycloud.ai/ my project: find, discover & explain AI research semantically https://findmypapers.ai/ My Patreon https://www.patreon.com/c/bycloud Can We Infer Confidential Properties of Training Data from LLMs? [Paper] https://arxiv.org/abs/2506.10364 Interpreting the Repeated Token Phenomenon in LLMs [Paper] https://arxiv.org/abs/2503.08908 Approximating Language Model Training Data from Weights [Paper] https://arxiv.org/abs/2506.15553 Try out my new fav place to learn how to code https://scrimba.com/?via=bycloudAI This video is supported by the kind Patrons & YouTube Members: 🙏Nous Research, Chris LeDoux, Ben Shaener, DX Research Group, Poof N' Inu, Andrew Lescelius, Deagan, Robert Zawiasa, Ryszard Warzocha, Tobe2d, Louis Muk, Akkusativ, Kevin Tai, Mark Buckler, NO U, Tony Jimenez, Ângelo Fonseca, jiye, Anushka, Asad Dhamani, Binnie Yiu, Calvin Yan, Clayton Ford, Diego Silva, Etrotta, Gonzalo Fidalgo, Handenon, Hector, Jake Disco very, Michael Brenner, Nilly K, OlegWock, Daddy Wen, Shuhong Chen, Sid_Cipher, Stefan Lorenz, Sup, tantan assawade, Thipok Tham, Thomas Di Martino, Thomas Lin, Richárd Nagyfi, Paperboy, mika, Leo, Berhane-Meskel, Kadhai Pesalam, mayssam, Bill Mangrum, nyaa, Toru Mon [Discord] https://discord.gg/NhJZGtH [Twitter] https://twitter.com/bycloudai [Patreon] https://www.patreon.com/bycloud [Business Inquiries] [email protected] [Profile & Banner Art] https://twitter.com/pygm7 [Video Editor] @Booga04