🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Using experience replay? Try ReLo! New work from our group 📢 Great job Shiva Somjit Nath Pedro Magalhães!

📢 New work with Somjit Nath on SSL for RL!

Work Done with Samira E. Kahou and @uboolean in collaboration with CM Labs - Vortex

Big thanks to my collaborators Shiva, Shruti Joshi, Vincent Michalski Danijar Hafner Samira E. Kahou! Happy to answer any questions! Paper: arxiv.org/abs/2210.11698 GitHub: github.com/arnavkj1995/VSG Drop by our poster on Tue Nov 29 11:00 AM -- 01:00 PM (CST) @ Hall J #409.

Exciting new work led by Ivaxi Sheth, will be presented today at the #NeurIPS2022 workshop on "Trustworthy and Socially Responsible Machine Learning" thanks to aamer AR Laya Rafiee Mohammad Havaei Mila - Institut québécois d'IA Imagia Canexia Health

📢 today at Deep RL Workshop #NeurIPS2022: Prioritizing Samples in Reinforcement Learning with Reducible Loss (arxiv.org/abs/2208.10483) Shiva Somjit Nath Pedro Magalhães and Locally Constrained Representations in Reinforcement Learning (arxiv.org/abs/2209.09441) Somjit Nath

Together with Ulrich Aïvodji and Marlos C. Machado, I am looking to recruit a postdoc for Fall'23 to work on: - Deep Reinforcement learning - Representation learning - Explainability - Privacy/Security See full description here: docs.google.com/document/d/1Jt… Mila - Institut québécois d'IA Amii École de technologie supérieure

Our new paper accepted at @midl-conference.bky.social led by Ivaxi Sheth @ NAACL 2025 📢

Our new paper accepted at #ICML2023, kudos to Somjit Nath Gopeshh Subbaraj and Khimya for the hard work!

Come see our poster today, Jun 19th at 12.30 pm PST, in West 215-216 of the Vancouver Convention Center! Sadly, we're not at CVPR, reach out to Nikunj Gupta for questions or comments. Thanks to CIFAR and Mila - Institut québécois d'IA

Very excited that our survey is public! A comprehensive review of transformers in reinforcement learning. Great work by Pranav agarwal, aamer AR, Pierre-Luc St-Charles and Simon Prince 🎉 CIFAR Mila - Institut québécois d'IA

Congratulations to Shiva Somjit Nath Pedro Magalhães, special thanks to all reviewers and AC for their constructive engagement in reviews and discussion period #NeurIPS2023 CIFAR Mila - Institut québécois d'IA

Happy to share that our paper "Combining modality specific deep neural networks for emotion recognition in video" has been selected as a Ten-Year Technical Impact Runner-Up at the 25th ACM ICMI Official 🎉 huge congrats to the team Mila - Institut québécois d'IA !

I will be in Toronto @ #MaRSMornings on Nov 9th discussing the potential of AI for disaster response. Thanks to MaRS and CIFAR for organizing this event.