AdvMLFrontiers

A timely reminder of our two rising star awardees who will be giving talks next week Saturday @NeurIPSConf , East Ballroom C, 15:30 and 15:45. Congratulations again!

Finally, we are glad to host Cornelia Caragea (NSF, US), who will be presenting , Saturday at 15:00: ‘Do LLMs Possess Emotional Intelligence?’. Thank you Cornelia!

Another highlight: @fraboeni at 10:10, presenting ‘From Risks to Resilience: Protecting Privacy in Adapted Language Models’. Thank you, Franziska!

We are also grateful for @Eleni30fillou‘s presentation about: ‘Unlearning memorized data from trained models’. Please join Saturday at 9:40 for this talk!

Further thanks go to Alvaro Velasquez (DARPA), who presents Saturday at 14:00 his talk ‘Frontiers in Adversarial AI: Autonomy and Biology’.

Another program sneak-peek: @HodaHeidari presenting, Saturday at 14:30, ‘Red-Teaming for Generative AI: Past, Present, and Future’. Thank you Hoda!

@maha81193 @lukgosch @NeurIPSConf Thank you for tagging @lukgosch! We couldn't find his profile when composing our post.

Our paper won the best paper award @AdvMLFrontiers #NeurIPS2024 🎉 Special thanks to my collaborator @lukgosch!🙂 We derive the first white-box robustness certificates for (graph) neural nets against data poisoning & backdoor attacks. Check out the talk on 14 Dec @NeurIPSConf

Program outlook: Thanks to @AlinaMOprea, who presents at 9:10 local time her talk 'Training Secure Agents: Is Reinforcement Learning Vulnerable to Poisoning Attacks?'

🏆 "Provable Robustness of (Graph) Neural Networks Against Data Poisoning and Backdoor Attacks" by Lukas Gosch, @maha81193, Debarghya Ghoshdastidar, @guennemann

🏆 "Can Watermarking Large Language Models Prevent Copyrighted Text Generation and Hide Training Data?" by @michael_panaite, @zorache_, @bang_an_, @Yuancheng_Xu0, Pankayaraj Pathmanathan, Souradip Chakraborty, @sichengzhuml, @tomgoldsteincs, @furongh

📢 We are delighted to reveal the Best Paper Awards: 🏆 Can Watermarking Large Language Models Prevent Copyrighted Text Generation and Hide Training Data? and 🏆 Provable Robustness of (Graph) Neural Networks Against Data Poisoning and Backdoor Attacks 👏👏👏

We look forward to your attendance and active participation in what promises to be an engaging and thought-provoking workshop on Dec. 14, 2024. 🌟

🚀 Not even a week left until our workshop @ NeurIPS 2024! 🚀 We’re delighted to share that the (tentative) program schedule for the AdvML-Frontiers workshop is now available! Check it out here: advml-frontier.github.io.

All of August left to prepare an amazing Workshop submission, allowing you to travel to Vancouver, discuss with the other #AdvML folks, and attend great invited talks! We are looking forward to your submissions!

For all young researchers, the AdvML Rising Star Award deadline is on September 2nd, Deadline for recommendation letters is September 9th. Please share with promising candidates!

📢 We're back with a new edition, this year at @NeurIPSConf in Vancouver! Paper deadline is August 30th, we are looking forward to your submissions!

Organizers are @sijialiu17 @pinyuchenTW @dongxiaozhu @RICEric22 @YaoQin_UCSB @KathrinGrosse @baharanm and @sanmikoyejo . We are looking forward to the workshop with all of you!

📢We would like to congratulate this year's winners of the best Paper award: 🎊 Cudos to @AlexRobey23, @fabianlat, @pappasg69, @HamedSHassani, and @CevherLIONS for their paper 'Adversarial Training Should Be Cast as a Non-Zero-Sum Game'! 🎉

📢 Our program is online: advml-frontier.github.io We are looking forward to receiving and discussing with all of you the new frontiers in adversarial ML!

📢And exactly one more week to go if you want to submit your awesome work! 👉New Deadline: May 28 '23 AoE