Kai Chen
@kaichen23
Ph.D. student @USCViterbi @USC_ISI. Computational Social Science & Natural Language Processing.
Excited for #EMNLP2024 in Miami 🌴next week! Join me on Nov 12 (Tue) from 16:00 to 17:30 in poster session 4 for this paper. Feel free to reach out!
🤔How susceptible are LLMs to Ideological Manipulation?⭕️ 🧐We find a concerning vulnerability: only a small amount of ideologically driven samples significantly alters the ideology of LLMs.🤖⚠️ 🔗arxiv.org/abs/2402.11725 @ZihaoHe95 @jun_yannn @taiwei_shi @KristinaLerman 🍀1/4
Our work "How Susceptible are Large Language Models to Ideological Manipulation?" was accepted by #EMNLP2024 Main Conference🥳. Huge thanks to my collaborators!
🤔How susceptible are LLMs to Ideological Manipulation?⭕️ 🧐We find a concerning vulnerability: only a small amount of ideologically driven samples significantly alters the ideology of LLMs.🤖⚠️ 🔗arxiv.org/abs/2402.11725 @ZihaoHe95 @jun_yannn @taiwei_shi @KristinaLerman 🍀1/4
Excited for #NAACL2024 in Mexico 🇲🇽 next week! Join me on June 19 from 11:00 AM to 12:30 PM in DON ALBERTO 1 for my talk on Safer-Instruct. Let's dive into alignment, synthetic data, and more!
Excited to get Safer-Instruct accepted to NAACL 2024 🥳! You don’t want to miss it if you want to reduce cost and boost efficiency in preference data acquisition 🚀. Check out our framework and dataset here: maksimstw.github.io/papers/saferin…
Honored to receive the 🏆 𝐛𝐞𝐬𝐭 𝐩𝐚𝐩𝐞𝐫 𝐫𝐮𝐧𝐧𝐞𝐫-𝐮𝐩 at the ICLR SeT LLM workshop! I will be giving a talk on this work on May 11th, 15:30, Schubert 6. Let's talk about AI Safety there! 🔐 Paper: arxiv.org/abs/2402.11725 Event: set-llm.github.io
🥳Exciting News! Our work, 🤖"How Susceptible are Large Language Models to Ideological Manipulation?" got 🏆𝐁𝐞𝐬𝐭 𝐏𝐚𝐩𝐞𝐫 𝐑𝐮𝐧𝐧𝐞𝐫-𝐮𝐩 at SET LLM #ICLR Workshop. Check our work here: arxiv.org/abs/2402.11725 Check the workshop here: set-llm.github.io
Excited to get Safer-Instruct accepted to NAACL 2024 🥳! You don’t want to miss it if you want to reduce cost and boost efficiency in preference data acquisition 🚀. Check out our framework and dataset here: maksimstw.github.io/papers/saferin…
🤔Enhancing LLM with RLHF is powerful, but ever wondered how to reduce costs and boost efficiency in preference data acquisition? 💰 🚀Introducing Safer-Instruct, a groundbreaking pipeline that complements humans to construct large-scale preference datasets efficiently. 🧵1/5
🤔Enhancing LLM with RLHF is powerful, but ever wondered how to reduce costs and boost efficiency in preference data acquisition? 💰 🚀Introducing Safer-Instruct, a groundbreaking pipeline that complements humans to construct large-scale preference datasets efficiently. 🧵1/5