Jonathan Lebensold (@jonlebensold) 's Twitter Profile
Jonathan Lebensold

@jonlebensold

RS @ Reliant. Researcher @ Mila / McGill.

ID: 21413440

calendar_today20-02-2009 17:08:51

788 Tweet

1,1K Followers

833 Following

André Picard (@picardonhealth) 's Twitter Profile Photo

Cybersecurity experts hail Quebec vaccine passport as model for the country. The #VaxiCode app, which scans data contained in the QR code, protects privacy and is almost impossible to fake, by Jacob Serebrin theglobeandmail.com/canada/article… via The Globe and Mail #COVID19 #vaccination #qcpoli

Steven Lachance (@lachances) 's Twitter Profile Photo

With Ontario following Quebec’s footsteps in rolling out vaccine passports, lots of Canadians have lots of questions around QR codes, security and privacy. I tried to provide a few answers on CBC News earlier today w/ Aarti Pole आरती. Part 1 of the interview 📺 👇

Steven Lachance (@lachances) 's Twitter Profile Photo

The #VerifyOntario iOS app includes a Google tracker that connects to their servers every time you open the app. Seemingly no personal information included, but very questionable practice. Quebec’s app does not do that. #COVID19ON #onpoli

Markus Anderljung (@manderljung) 's Twitter Profile Photo

Privacy-preserving machine learning – where ML systems can be trained on data while preserving data privacy – could allow users to only share their data for specific uses and could allow audits of developers' systems without access to proprietary or sensitive data.

Yu-Xiang Wang (@yuxiangw_cs) 's Twitter Profile Photo

Wondering the latest and greatest in privacy research for ML? The PriML'21 workshop #NeurIPS2021 will start in a few hours! priml2021.github.io There are activities in all time zones including 4 invited talks, 2 panels, 37 posters and 6 contributed talks. All are welcome!

Gautam Kamath (@thegautamkamath) 's Twitter Profile Photo

Very nice work on differentially private image classification. Thorough large-scale engineering with some new ideas (e.g., augmentation multiplicity) that seem to help a *lot*. Congrats to authors Soham De Leonard Berrada Hayes Samuel L Smith Borja Balle! arxiv.org/abs/2204.13650

Very nice work on differentially private image classification. Thorough large-scale engineering with some new ideas (e.g., augmentation multiplicity) that seem to help a *lot*. Congrats to authors <a href="/sohamde_/">Soham De</a> <a href="/LeonardBerrada/">Leonard Berrada</a> Hayes <a href="/SamuelMLSmith/">Samuel L Smith</a> <a href="/BorjaBalle/">Borja Balle</a>! arxiv.org/abs/2204.13650
Google DeepMind (@googledeepmind) 's Twitter Profile Photo

Training with differential privacy (DP) prevents models from leaking sensitive training data, but it often incurs a large drop in accuracy. In recent work, our team substantially improved the performance of image classification with DP. Read more: dpmd.ai/dm-dp-sgd 1/

Training with differential privacy (DP) prevents models from leaking sensitive training data, but it often incurs a large drop in accuracy. In recent work, our team substantially improved the performance of image classification with DP.

Read more: dpmd.ai/dm-dp-sgd 1/
Joey Bose (@bose_joey) 's Twitter Profile Photo

I'm on the job market for research scientist positions and post-docs! Most of my recent work has focused on geometric deep learning + generative models. If you think I'm a good fit for your lab/org do reach out. I'll also be at #NeurIPS2022 all next week and I'm happy to get ☕️.

Lucas Caccia (@lucaspcaccia) 's Twitter Profile Photo

I'm on the job market for research scientist & research engineer positions! My work revolves around transfer learning (continual, multitask). I will be at #NeurIPS2022, please reach out if you want to chat :)

AI at Meta (@aiatmeta) 's Twitter Profile Photo

📝 New from FAIR: An Introduction to Vision-Language Modeling. Vision-language models (VLMs) are an area of research that holds a lot of potential to change our interactions with technology, however there are many challenges in building these types of models. Together with a set

📝 New from FAIR: An Introduction to Vision-Language Modeling.

Vision-language models (VLMs) are an area of research that holds a lot of potential to change our interactions with technology, however there are many challenges in building these types of models. Together with a set
Reliant AI (@reliant_ai) 's Twitter Profile Photo

Thanks TechCrunch for covering our $11.3M seed round, bringing next gen(AI) analytics to biopharma and beyond. techcrunch.com/2024/08/20/rel… Happy to have great investors on board with Tola Capital, Inovia Capital and Mike Volpi in additon to our amazing Angels from before.

Marc G. Bellemare (@marcgbellemare) 's Twitter Profile Photo

With today's announcement Karl Moritz Hermann, Richard & I are thrilled to launch Reliant's next phase - building AI that will completely change how we work with data. Excited to bring Tola Capital, Inovia Capital, and Mike Volpi's expertise & experience on this journey. PS: We're hiring :)

Pablo Samuel Castro (@pcastr) 's Twitter Profile Photo

📢More world-class RL experts with startups that are hiring! I've worked very closely with Marc G. Bellemare , and would highly recommend it to anyone... apply to work with him and the rest of the Reliant AI team!

Haque Ishfaq (@haqueishfaq) 's Twitter Profile Photo

Excited to share the camera ready of our #ICLR2025 paper "Langevin Soft Actor-Critic: Efficient Exploration through Uncertainty-Driven Critic Learning". ➤ Camera ready: arxiv.org/abs/2501.17827 ➤ Code: github.com/hmishfaq/LSAC 🧵👇

Excited to share the camera ready of our #ICLR2025 paper "Langevin Soft Actor-Critic: Efficient Exploration through Uncertainty-Driven Critic Learning". 
➤ Camera ready: arxiv.org/abs/2501.17827
➤ Code: github.com/hmishfaq/LSAC 
🧵👇