Week Ending 9.27.2020
RESEARCH WATCH: 9.27.2020
This week was very active for "Computer Science - Artificial Intelligence", with 158 new papers.
The paper discussed most in the news over the past week was "The Radicalization Risks of GPT-3 and Advanced Neural Language Models" by Kris McGuffie et al (Sep 2020), which was referenced 10 times, including in the article Microsoft gets exclusive license for OpenAI’s GPT-3 language model in Venturebeat. The paper got social media traction with 9 shares. The researchers expand on their previous research of the potential for abuse of generative language models by assessing GPT-3. On Twitter, @AlexBNewhouse said "We now have an arxiv link for our paper on GPT-3! (This is my first arxiv submission, so I’m unreasonably excited even though it’s not that notable)".
The paper shared the most on social media this week is "Ethical Machine Learning in Health Care" by Irene Y. Chen et al (Sep 2020) with 247 shares. The researchers outline ethical considerations for equitable ML in the advancement of health care. @pranavrajpurkar (Pranav Rajpurkar) tweeted "Cool review paper! I especially like the highlighted five steps in ethical pipeline for healthcare model dev".
The most influential Twitter user discussing papers is Judea Pearl who shared "Exact parametric causal mediation analysis for a binary outcome with a binary mediator" by Marco Doretti et al (Nov 2018) and said: "A new mediation paper has crossed my screen: I was happy to see a parametric mediation formula for the direct and indirect effects on a odds ratio scale. Unhappy to see PO folks still referring to consistency as an "assumption" instead of "property."".
Over the past week, 199 new papers were published in "Computer Science - Computer Vision and Pattern Recognition".
The paper discussed most in the news over the past week was by a team at Tel Aviv University: "COVID-19 in CXR: from Detection and Severity Scoring to Patient Disease Monitoring" by Rula Amer et al (Aug 2020), which was referenced 10 times, including in the article RADLogics Announces FDA Clearance and Validation for AI-Powered... in PRWeb. The paper was shared 4 times in social media. The researchers estimate the severity of pneumonia in COVID-19 patients and conduct a longitudinal study of disease progression.
Leading researcher Luc Van Gool (Computer Vision Laboratory) came out with "Improving Point Cloud Semantic Segmentation by Learning 3D Object Proposal Generation".
The paper shared the most on social media this week is "Cloud Cover Nowcasting with Deep Learning" by Léa Berthomier et al (Sep 2020) with 161 shares. The authors focus on cloud cover nowcasting, which has various application areas such as satellite shots optimisation and photovoltaic energy production forecast.
The most influential Twitter user discussing papers is Judea Pearl who shared "Exact parametric causal mediation analysis for a binary outcome with a binary mediator" by Marco Doretti et al (Nov 2018)
This week was active for "Computer Science - Computers and Society", with 35 new papers.
The paper discussed most in the news over the past week was "The Radicalization Risks of GPT-3 and Advanced Neural Language Models" by Kris McGuffie et al (Sep 2020)
The paper shared the most on social media this week is "Ethical Machine Learning in Health Care" by Irene Y. Chen et al (Sep 2020)
The most influential Twitter user discussing papers is Judea Pearl who shared "Exact parametric causal mediation analysis for a binary outcome with a binary mediator" by Marco Doretti et al (Nov 2018)
This week was active for "Computer Science - Human-Computer Interaction", with 30 new papers.
The paper discussed most in the news over the past week was "PufferBot: Actuated Expandable Structures for Aerial Robots" by Hooman Hedayati et al (Aug 2020), which was referenced 3 times, including in the article PufferBot: A flying robot with an expandable body in Tech Xplore. The paper author, Hooman Hedayati (Researchers), was quoted saying "Our design offers several new communicative properties and potential safety benefits, as in the case of a collision or crash, the expandable structure can reduce impacts to both the robot and whatever the robot hits". The paper got social media traction with 5 shares.
This week was very active for "Computer Science - Learning", with 342 new papers.
The paper discussed most in the news over the past week was by a team at University of Minnesota: "The world as a neural network" by Vitaly Vanchurin (Aug 2020), which was referenced 14 times, including in the article Minnesota-Duluth professor says we may live in a neural network in City Pages. The paper author, Vitaly Vanchurin (University of Minnesota), was quoted saying "exhibit approximate behaviors". The paper got social media traction with 653 shares. On Twitter, @puishee_retina observed "I wonder where the inspiration comes from? Probably 2020 is just a bad local minima of SGD, maybe we should use ADAM instead? Oh wait, human are merely (or collectively as one) the hidden neurons of this giant neural network".
Leading researcher Kyunghyun Cho (New York University) published "SSMBA: Self-Supervised Manifold Based Data Augmentation for Improving Out-of-Domain Robustness" @BlancheMinerva tweeted "Very cool work, h/t".
The paper shared the most on social media this week is by a team at Google: "Tasks, stability, architecture, and compute: Training more effective learned optimizers, and using them to train themselves" by Luke Metz et al (Sep 2020) with 545 shares. The authors focus on general - purpose learned optimizers capable of training a wide variety of problems with no user - specified hyperparameters. @chipro (Chip Huyen) tweeted "One direction in AutoML I’m really excited about is learned optimizers: training optimizers to replace hand-designed optimizers (eg Adam, SGD). With 6k tasks and A LOT of compute, the authors found a learned optimizer that can train itself to be better 🤯".
The most influential Twitter user discussing papers is Judea Pearl who shared "Exact parametric causal mediation analysis for a binary outcome with a binary mediator" by Marco Doretti et al (Nov 2018)
Over the past week, 18 new papers were published in "Computer Science - Multiagent Systems".
Over the past week, 20 new papers were published in "Computer Science - Neural and Evolutionary Computing".
The paper shared the most on social media this week is by a team at Google: "Tasks, stability, architecture, and compute: Training more effective learned optimizers, and using them to train themselves" by Luke Metz et al (Sep 2020)
The most influential Twitter user discussing papers is Judea Pearl who shared "Exact parametric causal mediation analysis for a binary outcome with a binary mediator" by Marco Doretti et al (Nov 2018)
This week was active for "Computer Science - Robotics", with 52 new papers.
The paper discussed most in the news over the past week was "DroneLight: Drone Draws in the Air using Long Exposure Light Painting and ML" by Roman Ibrahimov et al (Jul 2020), which was referenced 4 times, including in the article Guiding light: New technology puts a light-painting drone at your fingertips in Tech Xplore. The paper author, Dzmitry Tsetserukou (Skolkovo Institute of Science and Technology (Skoltech)), was quoted saying "Flight control is a challenging task as user has to manipulate with the joystick to stabilize and navigate drones. Only a very skillful operator can maintain smooth trajectory, such as drawing a letter, and for the typical user it is almost not possible".