I managed to get in touch with Canada's foreign minister, and Yoshua Bengio and I explained the issues to her and her team. She has now helped get Canada's Minister for Immigration involved. Whilst there's no guarantee of good outcomes, this is good progress. Fingers crossed!

Hola AlRojoVivo de vez en cuando se nos cuela una falta de ortografía pero esto vuestro es: Cartel reaprovechado de los actos del 9N + edición con Paint + faltas de ortografía + no publicada oficialmente = que no sea una fake news de manual. #MásPeriodismo

Víctor Campos (Víctor Campos) from BSC-CNS Universitat Politècnica de Catalunya (UPC) presents his work on weight initialization for ResNets developed during his internship at Salesforce AI Research. Presented at #NeurIPS2019 #DLBCN arxiv.org/abs/1906.02341

New video from #DLBCN 2018 published: Víctor Campos (Víctor Campos) from BSC-CNS Universitat Politècnica de Catalunya (UPC) "Skip RNN: learning to skip state updates in recurrent neural networks", presented in #ICLR2018 youtu.be/SE2BknRa0t8

We present Explore, Discover and Learn (EDL), a novel paradigm for unsupervised skill discovery in RL. Work by Víctor Campos (Víctor Campos) from BSC-CNS. Co-authored with Alex Trott, Caiming Xiong Richard Socher Salesforce AI Research, & Jordi TORRES.AI BSC-CNS. arxiv.org/abs/2002.03647

CATALUNYA.-Barcelona Activa reconoce a Qbeast con un premio al mejor proyecto de alto impacto tecnológico. BSC-CNS lavanguardia.com/local/barcelon…

ICML 2020 (icml.cc/Conferences/20…) will be a virtual conference. We've been having discussions with ICLR folks about their plans, and will be learning from their experience (they are up first). We hope to enable as much of the normal ICML experience as possible, virtually.

Excited to introduce the AI Economist: Extends ideas from Reinforcement Learning for tackling inequality through learned tax policy design. The framework optimizes productivity and equality. Blog: blog.einstein.ai/the-ai-economi… Paper: arxiv.org/abs/2004.13332 Q&A: salesforce.com/company/news-p…

Interested in playing around with RL? We’re happy to announce the release of Acme, a light-weight framework for building and running novel RL algorithms. We also include a range of pre-built, state-of-the-art agents to get you started. Enjoy! bit.ly/2MlfeOA

During my PhD (In 2014), with Kyunghyun Cho I worked on un / self / semi-supervised learning of sentence representations. We evaluated these reps by training a *linear* classifier on low-data tasks. arxiv.org/abs/1602.03483 I always wondered if doing this made any sense? 1/N

Víctor Campos (Víctor Campos) will present "Explore, Discover and Learn (EDL)" for unsupervised RL at #ICML2020: 1) Tue Jul 14 11am. PDT / 8pm CEST 2) Tue Jul 14 11pm PDT / Wed July 15 8am CEST If you registered, find him at: icml.cc/virtual/2020/p… BSC-CNS Salesforce AI Research IDEAI-UPC Research Center

Two #internship opportunities to work with the Dolby Barcelona #AI team! Want to pursue research on learnable #speech representations for different downstream tasks? Let's do it under the supervision of Santi PdP and me. Check careers.dolby.com/job/Barcelona-…

SPREAD THE WORD: I'M HIRING PHD STUDENTS FOR FALL 2021 AT UPF Barcelona / @DTIC_UPF!! The project is about reinforcement learning theory and is funded by my recent European Research Council (ERC) Starting Grant. (#ERCStG) Details: cs.bme.hu/~gergo/jobs.ht… Deadline: Feb 12, 2021

A nice blog post on Seasonal Contrast is out 😀Oscar Mañas explains how to do self-supervised learning on satellite images by leveraging seasonal changes. Very simple and elegant implementation with @PyTorchLightnin devblog.pytorchlightning.ai/seasonal-contr…

Are you interested in open-ended learning? Reinforcement learning, multi-agent, neuroscience or learning dynamics design are close to your heart? Or maybe you have unique background/skills that you think are relevant? Lets make beautiful plots together deepmind.com/careers/jobs/2…!

If you are in Barcelona this December, you cannot miss the DL Bcn Symposium! Call for talks and posters is now open. Looking forward to it (my first on-site conference since pandemic🤯)

If you are interested in representation learning from a population of neural network models, have a look at our NeurIPS Conference paper today at the poster session. Konstantin @schurholt is happy to answer all of your questions - Xavi Giró SPCL@ETH @rengglic Víctor Campos

Thrilled to announce that "Human-level Atari 200x faster" has been accepted to ICLR 2026 ! Main contributions: - faster propagation of learning signals - handling differing value scales - efficient NN architecture - robust policy updates [paper]: openreview.net/pdf?id=JtC6yOH…

We’re excited to announce 𝗚𝗲𝗺𝗶𝗻𝗶: Google’s largest and most capable AI model. Built to be natively multimodal, it can understand and operate across text, code, audio, image and video - and achieves state-of-the-art performance across many tasks. 🧵 dpmd.ai/announcing-gem…

Today developers can start building with our first version of Gemini Pro through Google AI Studio at ai.google.dev.  Developers have a free quota and access to a full range of features including function calling, embeddings, semantic retrieval, custom knowledge