Week Ending 6.28.2020
RESEARCH WATCH: 6.28.2020
This week was active for "Computer Science - Artificial Intelligence", with 108 new papers.
The paper discussed most in the news over the past week was "Iterative Effect-Size Bias in Ridehailing: Measuring Social Bias in Dynamic Pricing of 100 Million Rides" by Akshat Pandey et al (Jun 2020), which was referenced 8 times, including in the article Uber and Lyft Respond to Algorithmic Bias Study Showing Price Increases for Travel to Non-White Neighborhoods in Complex. The paper author, Aylin Caliskan (George Washington University), was quoted saying "When machine learning is applied to social data, the algorithms learn the statistical regularities of the historical injustices and social biases embedded in these data sets". The paper got social media traction with 73 shares. The researchers develop a random - effects based metric for the analysis of social bias in supervised machine learning prediction models where model outputs depend on U.S. locations. A Twitter user, @DavidZipper, observed "Analyzing 100 million Chicago ride hail trips, researchers found significant evidence of bias. Algorithms used by Uber/Lyft/Via led to higher fares for those going to neighborhoods with a high share of minority or older residents, for example. DL link".
Leading researcher Yoshua Bengio (Université de Montréal) came out with "Rethinking Distributional Matching Based Domain Adaptation" The authors first systematically analyze the limitations of DM based methods, and then build new benchmarks with more realistic domain shifts to evaluate the well - accepted DM methods.
The paper shared the most on social media this week is by a team at Stanford University: "Compositional Explanations of Neurons" by Jesse Mu et al (Jun 2020) with 145 shares. @omarsar0 (elvis) tweeted "An approach for analyzing neurons compositionally to better characterize their behavior and obtain insights into model performance. It generates compositional explanations used to generate adversarial examples for manipulating model behavior".
The most influential Twitter user discussing papers is Sujoy Dhar who shared "Conflict in Africa during COVID-19: social distancing, food vulnerability and welfare response" by Roxana Gutiérrez-Romero (Jun 2020) and said: "A 10% increase in the local price index is associated with #sujoydhar".
This week was active for "Computer Science - Computer Vision and Pattern Recognition", with 283 new papers.
The paper discussed most in the news over the past week was by a team at University of California, Berkeley: "Motion2Vec: Semi-Supervised Representation Learning from Surgical Videos" by Ajay Kumar Tanwani et al (May 2020), which was referenced 18 times, including in the article Watch: AI and videos teach robot to suture in GlobalSpec. The paper got social media traction with 28 shares. The authors learn a motion - centric representation of surgical video demonstrations by grouping them into action segments/sub - goals/options in a semi - supervised manner. A user, @ShanthaRMohan, tweeted ""the team needed just 78 videos from the JIGSAWS database to train their AI to perform its task with 85.5 percent segmentation accuracy and an average 0.94 centimeter error in targeting accuracy."", while @EdwardDixon3 posted "A long way to go before it can pull on emergency torch-lit C-section, but really interesting to see this (simulated) stitching. Very label-efficient learning. Paper here: #IamIntel".
Leading researcher Yoshua Bengio (Université de Montréal) came out with "Image-to-image Mapping with Many Domains by Sparse Attribute Transfer".
The paper shared the most on social media this week is by a team at Toyota Research Institute: "Differentiable Rendering: A Survey" by Hiroharu Kato et al (Jun 2020) with 154 shares.
This week was very active for "Computer Science - Computers and Society", with 53 new papers.
The paper discussed most in the news over the past week was "Iterative Effect-Size Bias in Ridehailing: Measuring Social Bias in Dynamic Pricing of 100 Million Rides" by Akshat Pandey et al (Jun 2020)
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 by a team at Microsoft: "How good is good enough for COVID19 apps? The influence of benefits, accuracy, and privacy on willingness to adopt" by Gabriel Kaptchuk et al (May 2020), which was referenced 2 times, including in the article The app credibility gap in Nature Biotechnology. The paper got social media traction with 123 shares. On Twitter, @jsrailton observed "IMPORTANT STUDY: up to 50% of Americans wouldn't use a contact tracing app with privacy & accuracy problems. Research by: et al. Quick THREAD".
This week was extremely active for "Computer Science - Learning", with 575 new papers.
The paper discussed most in the news over the past week was "Iterative Effect-Size Bias in Ridehailing: Measuring Social Bias in Dynamic Pricing of 100 Million Rides" by Akshat Pandey et al (Jun 2020)
Leading researcher Yoshua Bengio (Université de Montréal) came out with "HNHN: Hypergraph Networks with Hyperedge Neurons".
The paper shared the most on social media this week is "Direct Feedback Alignment Scales to Modern Deep Learning Tasks and Architectures" by Julien Launay et al (Jun 2020) with 168 shares. @ChrSzegedy (Christian Szegedy) tweeted "Black magic. ;) My second attempt to try understanding this (last time a few years back). I still miss the intuition on how/why this works at all".
The most influential Twitter user discussing papers is Sujoy Dhar who shared "Conflict in Africa during COVID-19: social distancing, food vulnerability and welfare response" by Roxana Gutiérrez-Romero (Jun 2020)
Over the past week, 18 new papers were published in "Computer Science - Multiagent Systems".
The paper shared the most on social media this week is by a team at Cornell: "Stablecoins 2.0: Economic Foundations and Risk-based Models" by Ariah Klages-Mundt et al (Jun 2020) with 51 shares. The investigators seek to provide a sound foundation for stablecoin theory, with a risk - based functional characterization of the economic structure of stablecoins. @teo_leibowitz (Matteo Leibowitz) tweeted "in his new paper, provides framework for exploring these vectors. in some cases, results are unpleasant "In many DeFi systems designed with broad gov powers and without societal recourse, equilibrium participation of rational agents may be 0."".
The most influential Twitter user discussing papers is Sujoy Dhar who shared "Conflict in Africa during COVID-19: social distancing, food vulnerability and welfare response" by Roxana Gutiérrez-Romero (Jun 2020)
This week was active for "Computer Science - Neural and Evolutionary Computing", with 41 new papers.
The paper discussed most in the news over the past week was "The NetHack Learning Environment" by Heinrich Küttler et al (Jun 2020), which was referenced 4 times, including in the article Facebook releases AI development tool based on NetHack in Venturebeat. The paper got social media traction with 67 shares. The researchers present the NetHack Learning Environment (NLE), a scalable, procedurally generated, stochastic, rich, and challenging environment for RL research based on the popular single - player terminal - based roguelike game, NetHack. A Twitter user, @egrefen, said "We also provide nifty tools such as an agent analysis dashboard (with more coming). Contributions welcome! You can read more about NLE here: and download the env, tools, and agent code: Now go get that Amulet of Yendor! 🙂".
Leading researcher Pieter Abbeel (University of California, Berkeley) published "Locally Masked Convolution for Autoregressive Models".
The paper shared the most on social media this week is "Direct Feedback Alignment Scales to Modern Deep Learning Tasks and Architectures" by Julien Launay et al (Jun 2020)
The most influential Twitter user discussing papers is Sujoy Dhar who shared "Conflict in Africa during COVID-19: social distancing, food vulnerability and welfare response" by Roxana Gutiérrez-Romero (Jun 2020)
This week was active for "Computer Science - Robotics", with 53 new papers.
The paper discussed most in the news over the past week was by a team at University of California, Berkeley: "Motion2Vec: Semi-Supervised Representation Learning from Surgical Videos" by Ajay Kumar Tanwani et al (May 2020)
Leading researcher Ruslan Salakhutdinov (Carnegie Mellon University) published "Off-Dynamics Reinforcement Learning: Training for Transfer with Domain Classifiers" @towards_AI tweeted "Can we learn a policy for a target domain only using experience from a source domain with different dynamics? -- Modify the reward function so that the source domain resembles target domain: w/ B. Eysenbach, S. Asawa, S. Chaudhari".
The paper shared the most on social media this week is by a team at University of Oxford: "A Survey on Deep Learning for Localization and Mapping: Towards the Age of Spatial Machine Intelligence" by Changhao Chen et al (Jun 2020) with 94 shares. The investigators provide a comprehensive survey, and propose a new taxonomy on the existing approaches on localization and mapping using deep learning.