Mindblowing. Apple is now valued at a trillion dollars. CNN thinks other tech companies might not be far behind. I guess we are truly in the era of MAGA---Microsoft, Amazon, Google, Apple. money.cnn.com/2018/08/02/inv… Brown HCRI

This article speaks many, I believe, hidden truths about Quoc Le on GoogleBrain & seq2seq. Personally, I have enjoyed working with Quoc who cares less about credit assignment but rather teamwork and long-term vision :) medium.com/@aifrontiers/a…

Came across an old thread that seems to show folks struggle to pinpoint real-world success of contemporary RL (beyond game playing)

Accepted to #ICRA2019! One of the first instances of provably robust deep learning based controllers! We show Lyapunov stability while using a 4-layer deep neural net as part of the controller design. Experiments on a real robotic platform too.

This is a super cool resource: Papers With Code now includes 950+ ML tasks, 500+ evaluation tables (including SOTA results) and 8500+ papers with code. Probably the largest collection of NLP tasks I've seen including 140+ tasks and 100 datasets. paperswithcode.com/sota

I'm excited to announce that the 4th multi-disciplinary Reinforcement Learning and Decision Making conference (#RLDM2019) will be in Montreal this July 7-10! Check out the speaker list, and consider submitting an abstract (deadline March 1st)! Visit: rldm.org 🧠💻

Had a wonderful time at AAAI! Some notes: david-abel.github.io/notes/aaai_201… #AAAI19

“It is not that machines are going to replace chemists...It’s that the chemists who use machines will replace those that don’t.” Happy to see Mohammed AlQuraishi featured as well! nytimes.com/2019/02/05/tec…

Compelling essay: Roughly, (1) RL (like contextual bandits) is having a big impact on the real world, (2) That impact is not positive for humanity, (3) We as the engineers of these systems can do more to push for making future impacts positive. medium.com/@francois.chol… Brown HCRI

BBC News - Robot teaches itself to ice-skate bbc.com/news/av/techno…

Looks like a great overview of the area: "Algorithms for Verifying Deep Neural Networks," Liu et al.: arxiv.org/abs/1903.06758

Just posted "A Better Lesson", a short review of Ruch Sutton's recent post "The Bitter Lesson". rodneybrooks.com/a-better-lesso…

Impressive application of deep reinforcement learning to turbine optimization! Talk by Michael May from Siemens at #RVIAQC2019

Real-World Decision Making Workshop at #icml2019! realworld-sdm.github.io

thanks! here's the talk: slideslive.com/38916832/beyon… first paper mentioned is arxiv.org/abs/1803.00590 second paper will be on arxiv soon (will reply here when it's posted)

Notes from ICML Conference here: david-abel.github.io/notes/icml_201… #ICML2019

An amazing source of learning by Roger Grosse #metacademy metacademy.org/browse to learn new concepts in ML, in the depth you like. Love this effort!

Today I officially turn into a doctor (of philosophy). Special thanks to @YisongYue for being an amazing advisor. I have been fortunate to work with many awesome collaborators. And of course I would be no where without the support of my family and Zeynep <3

Policy Optimization with Linear Temporal Logic Constraints: arxiv.org/abs/2206.09546 1st author: Cameron Voloshin Co-authors: Swarat Chaudhuri & Hoang M. Le LTL can capture expressive constraints that are hard to do with reward engineering, such as an infinite loop (e.g. patrolling).

Go see our first 🤩 poster NeurIPS Conference tomorrow from Cameron Voloshin Hoang M. Le Swarat Chaudhuri Yisong Yue!! 👇👇👇👇👇👇