Michael S. Bernstein

Bio: Michael S. Bernstein is an academic researcher from Stanford University. The author has contributed to research in topics: Crowdsourcing & Computer science. The author has an hindex of 52, co-authored 191 publications receiving 42744 citations. Previous affiliations of Michael S. Bernstein include Association for Computing Machinery & Massachusetts Institute of Technology.

Ensemble: exploring complementary strengths of leaders and crowds in creative collaboration

Abstract: In story writing, the diverse perspectives of the crowd could support an author's search for the perfect character, setting, or plot. However, structuring crowd collaboration is challenging. Too little structure leads to unfocused, sprawling narratives, and too much structure stifles creativity. Motivated by the idea that individual creative leaders and the crowd have complementary creative strengths, we present an approach where a leader directs the high-level vision for a story and articulates creative constraints for the crowd. This approach is embodied in Ensemble, a novel collaborative story-writing platform. In a month-long short story competition, over one hundred volunteer users on the web started over fifty short stories using Ensemble. Leaders used the platform to direct collaborator work by establishing creative goals, and collaborators contributed meaningful, high-level ideas to stories through specific suggestions. This work suggests that asymmetric creative contributions may support a broad new class of creative collaborations.

Iris: A Conversational Agent for Complex Tasks

Abstract: Today, most conversational agents are limited to simple tasks supported by standalone commands, such as getting directions or scheduling an appointment. To support more complex tasks, agents must be able to generalize from and combine the commands they already understand. This paper presents a new approach to designing conversational agents inspired by linguistic theory, where agents can execute complex requests interactively by combining commands through nested conversations. We demonstrate this approach in Iris, an agent that can perform open-ended data science tasks such as lexical analysis and predictive modeling. To power Iris, we have created a domain-specific language that transforms Python functions into combinable automata and regulates their combinations through a type system. Running a user study to examine the strengths and limitations of our approach, we find that data scientists completed a modeling task 2.6 times faster with Iris than with Jupyter Notebook.

Personalization via friendsourcing

Abstract: When information is known only to friends in a social network, traditional crowdsourcing mechanisms struggle to motivate a large enough user population and to ensure accuracy of the collected information. We thus introduce friendsourcing, a form of crowdsourcing aimed at collecting accurate information available only to a small, socially-connected group of individuals. Our approach to friendsourcing is to design socially enjoyable interactions that produce the desired information as a side effect.We focus our analysis around Collabio, a novel social tagging game that we developed to encourage friends to tag one another within an online social network. Collabio encourages friends, family, and colleagues to generate useful information about each other. We describe the design space of incentives in social tagging games and evaluate our choices by a combination of usage log analysis and survey data. Data acquired via Collabio is typically accurate and augments tags that could have been found on Facebook or the Web. To complete the arc from data collection to application, we produce a trio of prototype applications to demonstrate how Collabio tags could be utilized: an aggregate tag cloud visualization, a personalized RSS feed, and a question and answer system. The social data powering these applications enables them to address needs previously difficult to support, such as question answering for topics comprehensible only to a few of a user's friends.

Scaling short-answer grading by combining peer assessment with algorithmic scoring

Abstract: Peer assessment helps students reflect and exposes them to different ideas. It scales assessment and allows large online classes to use open-ended assignments. However, it requires students to spend significant time grading. How can we lower this grading burden while maintaining quality? This paper integrates peer and machine grading to preserve the robustness of peer assessment and lower grading burden. In the identify-verify pattern, a grading algorithm first predicts a student grade and estimates confidence, which is used to estimate the number of peer raters required. Peers then identify key features of the answer using a rubric. Finally, other peers verify whether these feature labels were accurately applied. This pattern adjusts the number of peers that evaluate an answer based on algorithmic confidence and peer agreement. We evaluated this pattern with 1370 students in a large, online design class. With only 54% of the student grading time, the identify-verify pattern yields 80-90% of the accuracy obtained by taking the median of three peer scores, and provides more detailed feedback. A second experiment found that verification dramatically improves accuracy with more raters, with a 20% gain over the peer-median with four raters. However, verification also leads to lower initial trust in the grading system. The identify-verify pattern provides an example of how peer work and machine learning can combine to improve the learning experience.

On the Opportunities and Risks of Foundation Models.

Abstract: AI is undergoing a paradigm shift with the rise of models (e.g., BERT, DALL-E, GPT-3) that are trained on broad data at scale and are adaptable to a wide range of downstream tasks. We call these models foundation models to underscore their critically central yet incomplete character. This report provides a thorough account of the opportunities and risks of foundation models, ranging from their capabilities (e.g., language, vision, robotics, reasoning, human interaction) and technical principles(e.g., model architectures, training procedures, data, systems, security, evaluation, theory) to their applications (e.g., law, healthcare, education) and societal impact (e.g., inequity, misuse, economic and environmental impact, legal and ethical considerations). Though foundation models are based on standard deep learning and transfer learning, their scale results in new emergent capabilities,and their effectiveness across so many tasks incentivizes homogenization. Homogenization provides powerful leverage but demands caution, as the defects of the foundation model are inherited by all the adapted models downstream. Despite the impending widespread deployment of foundation models, we currently lack a clear understanding of how they work, when they fail, and what they are even capable of due to their emergent properties. To tackle these questions, we believe much of the critical research on foundation models will require deep interdisciplinary collaboration commensurate with their fundamentally sociotechnical nature.

Deep Residual Learning for Image Recognition

Abstract: Deeper neural networks are more difficult to train. We present a residual learning framework to ease the training of networks that are substantially deeper than those used previously. We explicitly reformulate the layers as learning residual functions with reference to the layer inputs, instead of learning unreferenced functions. We provide comprehensive empirical evidence showing that these residual networks are easier to optimize, and can gain accuracy from considerably increased depth. On the ImageNet dataset we evaluate residual nets with a depth of up to 152 layers—8× deeper than VGG nets [40] but still having lower complexity. An ensemble of these residual nets achieves 3.57% error on the ImageNet test set. This result won the 1st place on the ILSVRC 2015 classification task. We also present analysis on CIFAR-10 with 100 and 1000 layers. The depth of representations is of central importance for many visual recognition tasks. Solely due to our extremely deep representations, we obtain a 28% relative improvement on the COCO object detection dataset. Deep residual nets are foundations of our submissions to ILSVRC & COCO 2015 competitions1, where we also won the 1st places on the tasks of ImageNet detection, ImageNet localization, COCO detection, and COCO segmentation.

Very Deep Convolutional Networks for Large-Scale Image Recognition

Abstract: In this work we investigate the effect of the convolutional network depth on its accuracy in the large-scale image recognition setting. Our main contribution is a thorough evaluation of networks of increasing depth using an architecture with very small (3x3) convolution filters, which shows that a significant improvement on the prior-art configurations can be achieved by pushing the depth to 16-19 weight layers. These findings were the basis of our ImageNet Challenge 2014 submission, where our team secured the first and the second places in the localisation and classification tracks respectively. We also show that our representations generalise well to other datasets, where they achieve state-of-the-art results. We have made our two best-performing ConvNet models publicly available to facilitate further research on the use of deep visual representations in computer vision.

Very Deep Convolutional Networks for Large-Scale Image Recognition

Abstract: In this work we investigate the effect of the convolutional network depth on its accuracy in the large-scale image recognition setting. Our main contribution is a thorough evaluation of networks of increasing depth using an architecture with very small (3x3) convolution filters, which shows that a significant improvement on the prior-art configurations can be achieved by pushing the depth to 16-19 weight layers. These findings were the basis of our ImageNet Challenge 2014 submission, where our team secured the first and the second places in the localisation and classification tracks respectively. We also show that our representations generalise well to other datasets, where they achieve state-of-the-art results. We have made our two best-performing ConvNet models publicly available to facilitate further research on the use of deep visual representations in computer vision.

Deep Residual Learning for Image Recognition

Abstract: Deeper neural networks are more difficult to train. We present a residual learning framework to ease the training of networks that are substantially deeper than those used previously. We explicitly reformulate the layers as learning residual functions with reference to the layer inputs, instead of learning unreferenced functions. We provide comprehensive empirical evidence showing that these residual networks are easier to optimize, and can gain accuracy from considerably increased depth. On the ImageNet dataset we evaluate residual nets with a depth of up to 152 layers---8x deeper than VGG nets but still having lower complexity. An ensemble of these residual nets achieves 3.57% error on the ImageNet test set. This result won the 1st place on the ILSVRC 2015 classification task. We also present analysis on CIFAR-10 with 100 and 1000 layers. The depth of representations is of central importance for many visual recognition tasks. Solely due to our extremely deep representations, we obtain a 28% relative improvement on the COCO object detection dataset. Deep residual nets are foundations of our submissions to ILSVRC & COCO 2015 competitions, where we also won the 1st places on the tasks of ImageNet detection, ImageNet localization, COCO detection, and COCO segmentation.

Deep Learning

Abstract: Deep learning is a form of machine learning that enables computers to learn from experience and understand the world in terms of a hierarchy of concepts. Because the computer gathers knowledge from experience, there is no need for a human computer operator to formally specify all the knowledge that the computer needs. The hierarchy of concepts allows the computer to learn complicated concepts by building them out of simpler ones; a graph of these hierarchies would be many layers deep. This book introduces a broad range of topics in deep learning. The text offers mathematical and conceptual background, covering relevant concepts in linear algebra, probability theory and information theory, numerical computation, and machine learning. It describes deep learning techniques used by practitioners in industry, including deep feedforward networks, regularization, optimization algorithms, convolutional networks, sequence modeling, and practical methodology; and it surveys such applications as natural language processing, speech recognition, computer vision, online recommendation systems, bioinformatics, and videogames. Finally, the book offers research perspectives, covering such theoretical topics as linear factor models, autoencoders, representation learning, structured probabilistic models, Monte Carlo methods, the partition function, approximate inference, and deep generative models. Deep Learning can be used by undergraduate or graduate students planning careers in either industry or research, and by software engineers who want to begin using deep learning in their products or platforms. A website offers supplementary material for both readers and instructors.