Sam Roweis died unexpectedly on January 12, 2010.
He was a truly wonderful person; a beloved son, husband and father; and a treasured friend and colleague.
This is a place for all of us who were lucky enough to know Sam to share our memories and to help celebrate his life.
If you would like to add an article to this blog please contact email@example.com. Or you may leave a comment on any article. (Comments are moderated: please bear in mind that this is a place to remember Sam and to help celebrate his life.)
There is also an album of photographs for which contributions are welcome. Instructions on how to contribute appear next to album.
Tuesday, 1 February 2011
Steve Van Hooser
Tuesday, 2 November 2010
The following post is based on some reminiscences I shared during a memorial event held at the University of Toronto on January 22, 2010, and is recreated from my speaking notes. It’s been too easy to be “unable” to summon the energy/courage/right frame of mind to put these remarks in the form of a post, leading to many months of delay. - ceb
I’d like to say a few words about what Sam meant to the Department of Computer Science (DCS). As a friend, departmental colleague, and member of the same research group, Sam had a tremendous influence on me, just as he did on everyone he met. But as a Department Chair, I was privileged to know Sam on slightly different dimension than many of you. There are qualities that characterized Sam’s personal and scholarly life—his brilliance and creativity, enthusiasm and optimism, a genuine thirst for understanding and desire to change the world, and his generosity—he didn’t care who changed the world, or who got credit for it, he just wanted to contribute and help others contribute as well. He exemplified these same qualities as a departmental colleague and a member of the DCS family.
I first met Sam in 2001 when he interviewed here at the University of Toronto, and still recall the level of excitement that Sam generated, across the department and for me personally. I looked back last week through my email archive, and there is nobody in 19 years as I faculty member that I lobbied harder to attract than Sam (a lot of emails to Sam, to our chair at the time, and our hiring committee). We’ve interviewed some great candidates in this department, some of whom are much closer to me from a research perspective. But there was a spark in Sam that you knew was going to make his contributions to our research culture go far beyond those of someone with “mere scientific credentials.” He was clearly going to make the whole much greater than the sum of its parts. How true it was. I had just joined DCS a year and half earlier, and while I knew this was a terrific department, landing Sam that spring reinforced for me that this was a place on the move.
I wish I had more research interaction and collaboration with Sam. I became department chair only a couple of years after he arrived. This unfortunately limited my collaboration opportunities. Still he was one of those people who wanted to share everything. If you ran into him the hallway, he always had some tidbit of research to share. With Geoff, every couple of weeks he’d corner you as say, “I’ve finally figured out how the brain works. Do you have a couple minutes?” With Sam, it could be big stuff or small stuff. He was fascinated by it all. And it was rarely his own: I’ve got an email folder full of links from Sam saying “I read this paper, you might like it.” Or “Check out this cool Matlab code.” Or… It was simply part of his makeup, to share with his colleagues like this, whether inside and outside the AI group. Everything was fair game for him and that made him a terrific connector, something that every department needs.
Sam was a terrific supervisor with his own students and postdocs for the same reason. He believed that his role was to make everyone a better scholar. I’ve seen Sam in the classroom on a few occasions, and his talents as a teacher were off the charts. (I’ve also read hundreds of stellar teaching evaluations, and had to scramble to find lecture halls for his grad seminar, which routinely attracted 70, 80, 90 students and auditors from across campus). But what really left an impression on me was watching him interact with graduate students in research group meetings or thesis exams, including those of a number of my own PhD students. He would challenge them, push them to see extensions of their work, or connections with other work, or gradually lead them to find holes in their lines of argument themselves. But he was so gentle about it: if they didn’t understand a question or didn’t quite grasp what he was getting at, it was always “his fault” for not being clear enough, and he’d eventually just lead them until they reached their own conclusions. What could have been a completely intimidating experience was inevitably turned into an incredibly positive journey of discovery.
As a scientist, his own contributions were phenomenal. As a department chair, you dream of tenure cases like Sam’s, where the biggest challenge is finding enough superlatives to describe the quality and impact of his research, to describe the awards and recognitions, and explain his phenomenal teaching. The other challenge is finding referees—he worked with everyone! He was this way as a scientist. Josh Tenenbaum at MIT wrote on Sam’s memorial blog about Sam’s impact on the machine learning community, that to Sam “our field was one big joint effort.” He was there to help the entire machine learning community, and he was there to spread its impact far and wide. There aren’t a lot of computer scientists that publish in Science, in Cell, in the Astronomical Journal, in the Astrophysical Journal, and in the Journal of Computational Biology. And frankly I don’t think there are any, apart from Sam that published in them all.
He was the same way as a departmental citizen. He wanted to make our department a better place and wanted to help others do the same. Sam was not a huge fan of the pace at which we can effect change in a university environment: he was a man on the move with a lot to accomplish. But he knew that it mattered and he cared deeply. He approached DCS and his departmental colleagues (and department chairs) the same way he approached his science and wanted to help everyone be a bit better in what they do. And he’d occasionally have to treat a faculty member in a department meeting like a grad student, gently cajoling and leading them in a specific direction, but ultimately letting them draw their own conclusions.
Department chairs think often about the future of a department, about the next generation of leaders, and who has the respect and admiration of their colleagues, the personal qualities and commitment for leadership roles. Sam was someone whose name was always discussed in this way from the minute he arrived. But it was interesting with Sam: we had several conversations about this over the years, always at his instigation. He’d often say that he couldn’t understand how people like me (and Sven, vice-chair at the time) could devote themselves to an administrative job such as department chair. Sam had the leadership qualities to do it: frankly, he could do my job in heartbeat. You could tell he’d thought about it, and always had some worries about his abilities. But I think I understand now just what lay at the heart of his concerns. It comes down to Sam having to be fully invested, including emotionally, in anything he took on. As much as we all care about our colleagues and friends, about DCS and UofT, you sometimes have to detach yourself from the personal and emotional details of the job. I think Sam simply cared too much about his colleagues, about DCS, about anyone that he knew, to imagine how that was possible.
We talked a lot as he came to grips with the balance of family and career. When Sam and Meredith moved to San Francisco during his leave, and when they finally moved to New York, it was difficult for the department. But we all understood. And with Sam, we all knew he would remain part of the DCS extended family. Once you were in Sam’s orbit, you could not escape his gravitational field. In some ways it felt like we’d simply expanded as a department (as did NYU): and you could already see the stream of students and postdocs moving back and forth along the path that Sam helped build. It some sense it was like he hadn’t left at all. In early January, we had an extended email exchange about meetings and connections between DCS and Google. Sam’s departure from DCS in September was not a loss so much as it was a minor inconvenience in the grand scheme of things.
Now, we are without a doubt a much poorer department and university.
Sam, I’m sorry I never made it to the climbing gym with you; but Rich and I could never convince you to get out on the ice with us either.
We miss you more than mere words can express.
Tuesday, 27 July 2010
The breadth and spread of Sam Roweis' scientific legacy are just at their beginning. There are many people like myself out there who have just started to know him through his immense scientific work. As a newcomer to the field of dimensionality reduction, I eagerly read Sam’s paper on sensible PCA a couple of years ago and listed him, Zoubin Ghahramani, and Geoff Hinton as the people I would want to learn pattern recognition from. I admired his work without ever having had a chance to meet him in person or attend any of the conferences he likely attended or develop any form of personal acquaintance. My knowledge of him was through his work, which is imperishable. Today I had decided to do myself some good and read his 1999 paper on EM estimation of nonlinear dynamical systems and I came across in shock to the page with his obituary.
My message is only this: the legacy of his work is eternal and will grow further as years go by, well beyond the circle of colleagues who knew him in person. As this is the aspiration of any researcher, I guess this would have made him happy and I hope it will bring some consolation to his family. I do not sign this note as my name is insignificant and just stands for the many.
Thursday, 15 April 2010
Sam and Meredith stopped by my parent's house in Victoria, BC, when we were visiting, in the summer of 2006. They had just finished hiking the West Coast Trail, and had had a wonderful time. Many of the pictures from that trip made their way to their wedding web site. Meredith seemed such a perfect match for Sam, and they both seemed so fulfilled and full of hopes and dreams for their future life together.
Sam was instrumental in organizing the best wedding present I received. My wedding was to take place in Europe, where I had just moved, and I was bemoaning the fact that many of my North American friends were not going to come, including Meagan, one of my best friends from undergrad. Not long before the wedding I received an email from Sam, asking if he could bring someone he had just met, who “curiously” was also named Meagan. I thought nothing of this coincidence and burst into tears of joy when I picked them up at the train station.
I knew Sam best when we were both in grad school. We are not in the same field, so I was only indirectly aware of his academic brilliance. Besides being one of my closest friends, he was my coach; I taped an email from him to my shelf so that I would see it every time I looked up from my desk when studying. I have a T-shirt from him with the “words of wisdom” of the poster on his wall. I have reproduced both below.
It has “already” been over three months since one earthquake shook Haiti and another shook the world of Sam's family and friends. I still think of Sam (and Meredith, Aya and Orli) often. Sam will be remembered, for so many great reasons, by so many people, for a long long time to come.
I say never let go, never rest, never stop, never give in, never show fear, never tire. Cleaner, faster, better than everybody else, more practice, later nights, earlier mornings. Only the mediocre are always at their best. Push yourself. Keep fire in your eyes and fire in your heart and always do what you think is the right thing.
Saturday, 10 April 2010
Sam went to the famous high school (University of Toronto School) along with several of our other friends. For a while Sam was dating an engineering girl Rina, so I'd see him on the girls' side of the residence. But more often, I'd see him when I went over to the guys' side. We would congregate in someone's room, talking about anything, everything, for hours. Some of us huddled on the tiny bed, some spread out on the floor, some sat on the desk. These long talks fascinated me. They opened my mind to so many things I had never heard about.
One day, Sam talked about an experiment he did in high school. He wanted to know if the mind failed earlier than the body. He and 2 high school friends stayed up for 3 days with no sleep; they talked, did quizzes and rode a stationary bicycle. They were ok on day 1 and 2, but by day 3, they could still pedal the bike, but the blur coming out of their mouths no longer made sense. Sam always had an inquisitive mind.
I last saw Sam in 2004 at a small reunion. He looked the same. Ten years hadn't changed him a bit. The photo on this memorial site was taken at our reunion, very familiar and shocking when I saw it. At the time, he was consulting for Google, and very proud of it. I have always liked Sam.
Tuesday, 6 April 2010
I was a fellow Shad-ite in exactly the same program as Sam (Waterloo 1989). I was shocked and saddened to learn of his unexpected passing through the Shad Valley website.
I have wonderful memories of Sam from that summer. His sense of humour was unique, and his intellect dazzled us all. He was a brilliant and original thinker who blazed his own path. He was older than his years in maturity, and had a compassionate side that was rare for people of our age group. Though I didn't keep in touch with him after our 4 weeks together in the halls of Conrad Grebel and the UW campus, I am not surprised that he led an unparalleled career in academia before his unfortunate and untimely death.
As a parent, my heart goes out to his twin daughters, and I hope that they and their mother will be surrounded by much love, courage and support as they carry on without their beloved Sam. I cannot imagine what Sam's wife must be going through, but I hope that the memories that I and many others share of Sam will be of some comfort. That he lives on in our collective memories. To know that we will always remember him, and in that respect we keep a bit of him alive.
89W Shad participant
now a stay-at-home mom of 3 somewhere in Montreal, Quebec, Canada
Sunday, 7 February 2010
Friday, 5 February 2010
Sam Roweis (1972-2010)
Scientist and Engineer
The Department of Computer Science, the Courant Institute, and New York University mourn the untimely death of Professor Sam T. Roweis, who took his own life on January 12, 2010. Sam was a brilliant scientist and engineer whose work deeply influenced the fields of artificial intelligence, machine learning, applied mathematics, neural computation, and observational science. He was also a strong advocate for the use of machine learning and computational statistics for scientific data analysis and discovery.
Sam T. Roweis was born on April 27, 1972. He graduated from secondary school as valedictorian of the University of Toronto Schools in 1990, and obtained a bachelor's degree with honours from the University of Toronto Engineering Science Program four years later. His first exposure to AI and neural computation occured when--as an exceptional undergraduate--he took the graduate-level Neural Network course taught by Geoffrey Hinton. Here Sam discovered what would become his lifelong interest: unlocking the mysteries of intelligence; motivating all his work was a dream to understand human intelligence, and to build intelligent machines.
In 1994 he joined the Computation and Neural Systems PhD program at the California Institute of Technology, working under the supervision of John J. Hopfield. Sam made several contributions to the then-nascent field of molecular and DNA computing. With his contemporary Erik Winfree and others, he made a proposal for a sticker-based model of computation. But the main topic of his thesis was speech recognition, time-series analysis, and dynamical systems modeling. A central theme of his research was the systematic use of probabilistic frameworks to formulate and analyze learning algorithms. He realized that non-linear dynamical systems could be learned using the Expectation-Maximization (EM) algorithm. He proposed a variation of the well-established Hidden Markov Model (HMM) method for speech recognition, and he showed how a new form of Independent Component Analysis (ICA) could be used to separate multiple audio sources from a single microphone signal. He also realized that Principal Components Analysis (PCA) could be re-interpreted as the limit of a probabilistic model. His PhD research culminated with the publication of a landmark 1999 article, co-authored with Zoubin Ghahramani, that demonstrated that HMM, ICA, PCA, and Kalman Filters can all be seen as variations on a single linear Gaussian model.
After earning his PhD in 1999, Sam took a postdoctoral position in London with the Gatsby Computational Neuroscience Unit founded by Geoff Hinton. Sam's enthusiasm and creativity played an important role in making the Gatsby Unit one of the top labs in computational neuroscience. At Gatsby, Sam started an incredibly fruitful long-distance collaboration with Lawrence Saul (then at AT&T Labs in New Jersey), which led to the Locally Linear Embedding algorithm (LLE). The LLE paper, published in Science in 2000, revolutionized the field of dimensionality reduction, and gathered over 2700 citations in less than 10 years. It spurred an entire new sub-field of machine learning, called manifold learning, and gathered a considerable amount of interest from other technical fields, including applied mathematics. With LLE, Sam and Lawrence taught us to "think globally and fit locally": Given points in a high-dimensional space, local geometric relations among groups of nearby data points capture both local and global structure in the whole data set. This permits organization, visualization, and search for large, complex data collections. The method has had numerous applications in data visualization for biology, neuroscience, and the social sciences.
After his postdoc, some time at MIT, and a stint with the startup company WhizBang! Labs, Sam took a faculty job at the University of Toronto, to which Geoff Hinton had returned. In making this choice Sam rejected several extremely prestigious offers for the unparalleled intellectual atmosphere he found at Toronto surrounding his mentor, Geoff.
In this period, two new unsupervised methods he developed were Stochastic Neighborhood Embedding (SNE) and Neighborhood Component Analysis (NCA). Both methods use the idea of learning a function that maps datapoints into a space in which semantically similar objects are nearby, while semantically dissimilar objects are far apart. SNE has become a popular method for visualizing and organizing high-dimensional data, while NCA has spurred a resurgence of interest in methods that learn similarity metrics from data. Sam published a set of papers on speech and signal analysis, particularly using factorial HMM and hierarchical models. He was appointed to a Canadian Research Chair in statistical machine learning, and received a Sloan research followship in 2004.
In 2005 Sam spent a semester at MIT. Capitalizing on his work on blind source separation, he co-authored a landmark 2006 SIGGRAPH paper with Rob Fergus and others on removing camera shake from a single photograph. It was during his stay in Cambridge, MA that he met his wife, Meredith Goldwasser.
While at MIT and upon his return to Toronto, he focused on using machine learning and statistical methods to contribute to other sciences, such as astronomy and biology. He started an extremely fruitful collaboration with NYU astronomer and secondary-school friend David W. Hogg. Their most visible success together was a kind of search engine for the sky, called Astrometry.net. The system can take any picture of the sky from any source, and instantly identify the location, orientation, and magnification of the image, as well as name each object (star, galaxy, nebula) it contains. Sam and David introduced astronomers to a number of large-scale statistical methods that enabled increasingly automated and precise data analysis. One of their methods can even estimate the year at which an image was taken by measuring tiny variations in stellar positions.
In 2006 he was named a fellow of the Canadian Institute for Advanced Research (CIfAR) and received tenure at Toronto. Sam was not just a scientist, however, he was also an engineer. When Meredith took a job at Genentech in San Francisco, Sam took an opportunity to have a more direct impact on the world by joining Google's research labs in San Francisco and Mountain View in 2007. He was fond of saying that Google's search engine is one of the closest things we have to an intelligent computer.
In the summer of 2008, Sam and Meredith's twin daughters, Aya and Orli, were born in San Francisco. They were born very prematurely and had to be kept in intensive care unit for many weeks. Sam took an extended paternity leave from Google to take care of the twins.
Sam's stay at Google, and the success of his computational astronomy work, had renewed his interest in academic research. He decided to join the Computer Science Department at NYU's Courant Institute as an Associate Professor, and moved the family to New York City in September 2009. At NYU, his collaboration with David Hogg redoubled, as did his ongoing collaborations with Rob Fergus and Yann LeCun. His passing leaves many open threads, and many projects unfinished; at the time of his death he was working on beautiful, simple, but game-changing ideas for astronomical data analysis and remote sensing.
Sam had a singular gift: to him, any complex concept was naturally reduced to a simple set of ideas, each of which had clear analogies in other (often very distant) realms. This gift allowed him to explain the key idea behind anything in just a few minutes. Combined with contagious enthusiasm, this made him an unusually gifted teacher and speaker. His talks and discussions were clear and highly entertaining. His tutorial lectures on graphical models and metric learning, available on video at videolectures.net, have been viewed over 25,000 times. He would often begin group meetings by giving a puzzle, the solution of which was always beautiful, enlightening, or hilarious.
Many members of the research community became friends with Sam, because of his warm and friendly personality, his communicative smile, and his natural inclination to engagement and enthusiasm. Sam inspired many students to pursue a career in research, and to focus their research on machine learning and artificial intelligence. Already in his short time at NYU, Sam had become a key member of the computer science department, thanks to his broad interests, his clear-sightedness, his sense of humor, his warmth and his infectious enthusiasm. He was also a loving and devoted father to the twins and husband to Meredith.
In the last year of his life, he was battling hopelessness and pessimism about his life and career. His wonderful professional demeanor concealed these problems from most his friends and colleagues. He is greatly mourned by his colleagues and students at NYU, who extend their sympathy to his many friends in the broader research community, especially at the University of Toronto, the Gatsby Neuroscience Unit, and Google Research. Most of all, we express our deepest sympathy to his wife Meredith, his twin baby daughters Aya and Orli, and his father Shoukry.
-- Yann LeCun, David Hogg, Zoubin Ghahramani, Geoffrey Hinton.
Sam Roweis's home page and publications
Wednesday, 3 February 2010
Monday, 1 February 2010
Sunday, 31 January 2010
I remember Sam for his boundless energy, infectious enthusiasm, and generosity. One of my favourite memories of him is of several discussion sessions at the whiteboards in the tearoom at the Gatsby during his stay there 1999-2001, in particular with Geoff Hinton, Peter Dayan and Zoubin Ghahramani. These were a lot of fun, exploring ideas with energy, curiosity and a pleasure in finding things out.
It was always a treat to meet Sam at a conference or on a research visit. He would always have new things to tell, whether about his own work, or neat things that others had done. For example I remember him standing in my old office in Forrest Hill, Edinburgh, telling me about a backward pass algorithm for sampling from the posterior distribution in HMMs, and how it was a nice NIPS-y generalization of the Viterbi algorithm.
I last saw Sam in early December 2009 at a meeting in Vancouver. It makes me very sad to think I won't see him again, but perhaps one way for us to remember him is to try to aspire in some part to his standards of enthusiasm, creativity and curiosity.
Very early in the morning after his death, I awoke with a message from Yann LeCun about the news. I was completely shocked. I felt disbelief and a loss akin to one I had felt a few years earlier when my closest cousin had died in a meaningless accident. Writing these words my throat is bursting with emotion. I am clinging to the image of his communicative smile.
So long, Sam, and thank you for everything you brought to this world. I will forever remember the beautiful human being you were, and for me and the many others you have interacted with, you will remain an intellectual giant of our community and hold a special place in our heart.
Monday, 25 January 2010
Since I heard the awful news, I'm sure like many others I've repeatedly gone through my email, my Sam folder, dating back to when he first arrived, in 2001, a year after I had come to Toronto. His Science paper was just published, and he bowled everyone over during his visit, both in content and style.
Yet he sought my advice quite a bit at first, about funding, students, teaching, grants. Very soon it became clear to me that he needed no help from me, because he excelled in everything right away, as a:
- Faculty member: He threw himself into every issue and discussion, and we could always count on him for the pithiest, and often funniest comment at the meeting.
- Department citizen: He was some needed glue in the department, drawing it together. He regularly poked his head into faculty offices and attended meetings in other research groups.
- Great friend, and personal booster. I still remember a distinguished visiting speaker showing up for his allotted time-slot with me, which I was nervous about, yet he came in very excited about my newest project. I couldn't figure out how he knew, let alone why he was excited, until I realized that he had just talked to Sam, who had used some of his time telling the guy about it!
- Genuinely nice guy: It is the little things really show someone's character. With Sam one thing I recall is that early on I let him use my office and computer. I had a coral reef background, with tropical fish swimming. When I returned, there was a battery-powered water-filled coral reef toy, complete with swimming fish, next to my monitor. It is still there today.
After that I watched his star steadily rise. Yet his head did not swell: he still took an interest in everything, and was helpful to everyone. And he was a ton of fun at parties -- he somehow managed to be the goofiest and funniest guy at late-night parties at the conferences, without having touched a drop of alcohol.
He was also a ton of fun to talk research with, and to collaborate with on projects: a fountain of good ideas, generous with his time, and very respectful of student's contributions. I can remember several times, where we'd be stuck, in the depths of some derivation or failing simulation. And Sam would jump up, bounce over to the whiteboard, and write down a simple insight or equation that cut through the mess, putting us on track.
Finally, there are a few lessons I've learned from him I'm determined to try to emulate: his clear thinking, his lucid and simple explanations, but also his personal style, his genuine pleasure in others' successes. And he has also made me value my work friendships all the more -- realizing how important they are, and also what a great set of people there are out there.
Saturday, 23 January 2010
Several of those who have contributed to the blog have noted Sam’s talent as a teacher. Even as a little boy he demonstrated that talent. After I finished grad school I worked for a year for Sam’s late mother Nora who had a consulting firm – International Development Organization (IDO). Sam, then 13 years old, would come after school to IDO to help me with computer-related things like how to change the printer cartridge. He insisted that he would teach me how rather than simply doing it himself. In particular I remember the printer cartridge lesson given sweetly and good naturedly by him without a hint of condescension.
I moved back to Winnipeg to accept a job and during the many intervening years we lived our lives going on our separate journeys with our paths rarely crossing. One of the most notable memories I have of Sam as an adult is his wedding which stands out in my mind as the most enchanting one I have ever had occasion to attend! Initially I was apprehensive about going because apart from my uncle Shoukry and his partner Heather I did not know the other guests and wrongly assumed that with my liberal arts background Sam’s scientist friends would converse on topics that are completely inaccessible to me who has never figured out how to enter phone numbers of contacts on her cell phone and instead carries around a small address book in her handbag)!
Most thankfully I set aside those apprehensions and went and spent the week-end in Jacob, Ontario with a close-knit group of Sam and Meredith’s dearest friends and I remember thinking during the wedding event that there was just so much love in the room.
One definitely learns about a person through the company they keep. I was so glad to have had the opportunity at Sam’s wedding to meet and spend some bit of time interacting with Meredith as well as Sam and Meredith’s friends who struck me as among the finest people quite apart from their scientific or professional stature. They came across as interesting and more importantly interested, articulate, warm and genuine. Through my participation in his wedding, I came to learn how much my cousin Sam valued friendships.
While he and I did not have occasion to discuss them, in contrast to our other cousins, Sam and I shared certain life-defining experiences in common: we are both only children, we were both born in countries outside Egypt and both of our mothers died when they were comparatively young.
Arriving at a place of acceptance of events that we find overwhelming to fathom, and cessation of the searching mind that is hungry for answers, are entirely understandable challenges at this time; and yet, for Sam’s sake, I believe the greatest gift we can give him now is to release him in peace unconditionally, and to cherish the memories we have of our times together with him. How wonderful indeed that in such a relatively short life such as his, Sam managed to cast such a wide net in terms of his personal and professional impact. It leaves me very proud of him.
Thursday, 21 January 2010
Meredith, Sam loved you with his entire huge heart, often calling you Mer with a sparkle in his eyes. When Dahlia and I first inquired about you, Sam proudly said that you have a Ph.D. in biostatistics. His lofty tone when he answered our questions about you stood in contrast to the humble tone he used when speaking about his own research. Indeed, he was low key when referring to himself, nonetheless, he was also the most animated and colorful researcher I have ever met. Soon after you and Sam settled in the Bay Area, I noticed that Sam often arrived rather late to Google. I had to use my interrogation skills in order to pull out minimal details. Sam disclosed that he had to spend a morning, here and there, with you at a doctor's office. Still, I did not hear a single complaint, and he always had an admiring and compassionate tone when mentioning you. He happily and wholeheartedly adopted your heritage, lighting Hanuka candles with us, celebrating Purim. Dahlia and I jokingly called Sam our honorary Jew. Sam was clearly the man of all religions, accepting and acknowledging any kind of spiritual ritual.
Aya and Orli, Sam was a fantastic father, loving, caring, warm, and a great teacher. You are too little to understand but you have inherited his enthusiasm and joy. Sam's family and friends will be next to you to help you grow into wonderful kids and joyful teens.
Shoukry, we have never met until the sad weekend, but I heard so much about you from Sam. Sam was no doubt your flesh and blood, a young version of your creative spirit. A while ago, Dahlia and I visited Meredith and Sam, and Sam told us a great story about you. You wanted to make a special dish but lacked the right pot for cooking. You thus went to the basement, where you keep all kind of tools, and spent a couple of hours building the perfect pot, rather than going to the store. The pot was then used once or twice... After hearing the story Dahlia and I had a big smile, and I said, "Sam, aren't you pretty much the same? Don't you have a similar mindset when you conduct your own research?". Shoukry, Sam loved you and admired you. How I wish he was less of a perfectionist.
As I wrote, I will leave it to others to write about Sam's brilliance and scientific achievements. Let me tell you a bit more about Sam's huge heart. Months ago, when it was clear that Sam was heading back to academia, Sam, Mark, and I had lunch together at one of Google's cafes. Mark was prickly as he was struggling with his project at Google. I was bitter because my initiatives seemed to reach a dead end. Sam was Sam, all positive. Do not get me wrong. Sam had his own share of frustration at Google. He designed, implemented, and (almost) launched a beautiful project only to see it hit a glass ceiling. During that lunch, I was pointing to all sort of faults and quirks while Sam was able to find a kind and positive angle to any issue I surfaced. Finally, Mark said, "Leave the kid alone, he is a genuine, sincere, goody two-shoes."
Let me finish with a true tale that symbolizes Sam to me. It was Google's "Bike to Work" day. Sam was riding his bike from San Francisco to Mountain View with a bunch of young engineers from Google, quite a feat. I saw Sam in the middle ofthe group of sweaty cyclists at the line to breakfast. I was post run, clean, dry, in long pants and a long sleeve shirt. The youngsters looked at me with a somewhat condescending look. Sam noticed it right away and said with a grin,"Look guys, Yoram likes to bend the rules, he 'Ran to Work'".
Sam and I first met when I joined the Gatsby Unit as a grad student back in 2000, and I was immediately struck by his tremendous friendliness, humour and warmth -- he was just great at making people feel naturally welcome and at ease. I have vivid memories of him presenting machine learning courses, and many tea talks and journal clubs whilst we were there. He had an uncanny ability to take complicated ideas, and convey them (excitedly) in such a way as to make them seem natural (and frequently funny). An iconic recollection of Sam is the way he'd come up to you with a gleam in his eye and tell you about the latest cool algorithmic trick he'd found out about; always genuinely happy and excited to share his knowledge. Then he'd proceed to explain the details lucidly and concisely -- often taking a whole paper or subject area and condensing it into a key insight so that you just "got it".
There are also loads of great memories from the many brunches, dinners, parties and random social occasions in London, Toronto, and of course NIPS. Consistently good times. And Sam would always make sure that everyone present felt welcome and included. He was as fun to be around socially as he was rewarding to interact with professionally -- which is to say extremely.
To Meredith and his family, my heart goes out to you. I find the events incredibly hard to process and so can only begin to imagine the pain of your loss.
Sam, you'll be missed dearly and fondly remembered. The personal and intellectual impressions you left on me, and on so many others, will ensure that even in death you'll live on in our minds, our work, and in our hearts.
Wednesday, 20 January 2010
I have been trying to remember the first time I met Sam, and I think it must be when he and Aaron came roller skating past my front door just after they’d arrived at Caltech. They were on a mission either/both to map the topology of the graduate student apartment pathways and to make friends. Or maybe it was to find out where all the single women lived, they were pretty cagey about their purposes. Whatever the intent, I was charmed and happy that Sam’s fun bus had arrived.
Sam was great at fun. I hope that David Hogg gets to chronicle here their determinedly legendary parties at Princeton. I think they tried to get every single ambulatory person in New England to attend, and even set up a travel fund to entice the more distant potential partiers. They “imported” turntables from New York, and Sam impressed with his B-Boy moves, which I assert here were awesome, and not just awesome for a scientist.
Though we were friends at Caltech, it was the few years after that I got to know Sam a lot better, as we ended up in the same cities for periods of time and went through some of the big life changes at the same time. Sam was lovely to have as a friend. As has been said here so much, he cast a glow and made you feel better and special, and somehow encouraged one to _be_ better and more special. I am grateful for the years that we were close.
Things turned out well with Sam around. When I moved to Cambridge (MA) Sam was going to help me start to get to know my city. We settled on the approach of just skipping the maps and getting on some buses and start taking them around Boston in our own random walk. This appears to be out of character for Sam who seems in other posts here to have “optimal strategies,” but that’s what we did. We ended up at a dusty fantastic Diagon Alley kind of bookstore full of old maps and communist propaganda posters, and I remember Sam getting totally engaged with the owner over some discovery he made and buying presents for people and promising we’d come back if we could figure out how we’d gotten there in the first place. Ten years later I still have no idea.
But he was so attentive to details, he probably did remember. He diligently made the effort to do the small things that build an intimacy and friendship, things most of us would forget to do as the moment passed. My friend Diana and I had him over for dinner once, and he was apparently delighted at my fake French accent reading of a Harper’s magazine article about Brigitte Bardot—he made me feel like I was way funnier than I actually am—and shortly after, a subscription to Harper’s mysteriously began arriving, followed by a note from Sam urging me to keep up my dramatic recitations. I met Diana for a drink after Sam’s memorial yesterday, and she reminded me about when I sent her (having never met him) to stay with him in London, and she recalled his excellent hospitality to a stranger; for example, he prepared her a choice of several colors for her “guest toothbrush.”
It was a couple of years later when Sam came back to spend some time at MIT, and he and I both were asking the same kind of big life questions. Sam was searching deeply to make the right decision about whether to take an enormously prestigious and demanding job at MIT or to return to Toronto, to probably a less stressful job where he could be near his family and friends. I was selfish at wanting him to stay in Cambridge and promised to fish him out of the lab and keep his social life going, but in truth I was really proud of him for making a choice that asserted value for both his personal and professional selves. I thought I shouldn’t worry about him because he was not just smart but wise.
I was weighing the same kind of decision about leaving academia to work on issues that deeply compelled me but are less valued professionally. He encouraged me to make this leap and believed that I would do good in the world this way.
Despite all of the other demands on him, personal and professional, and the ways of peripatetic scientists, he endeavored to maintain the threads of connection. It has been a few years since we’ve seen each other, but Sam was brilliant at remembering every year to say happy birthday, or just how does the day find you, when are we going to hang out? I remember when he wrote to tell me he was engaged and blissed out; strangely, it was the same day I was writing to invite him to my own wedding, though as it turned out Ahmed and I chose the same day to get married that Meredith was moving up to Toronto. Though we never got to meet each other’s “missing piece,” I was grateful that he had found what he was looking for, and continued to make steps toward a fulfilled life.
As so many have mentioned, Sam was so upbeat and positive, one might assume he was that way because he was so incredibly gifted that life was easy, but he like us all was not a stranger to sadness and uncertainty and the knowledge that we would at least occasionally fail as human beings. I think his joie de vivre and warmth and openness were a deliberate and brave and uncommon choice. Of all his considerable talents, that is the aspect of Sam I admired most.
I have been trying to sort out why my own grief at this news of Sam has been so overwhelming. Part of it is feeling the loss of someone so extraordinary and special. But also so painful is knowing that our friend’s awesome abilities to be positive somehow became depleted enough that for a moment he wasn’t able to extend the same kindness to himself that he unfailingly extended to us.
I hope that Sam’s friends and family, and especially his father and Meredith and the girls, find peace and I am sending my best wishes for that and for the girls to be as joyful in life as he was.
Sam, you were dear to me!
I interacted with Sam a lot while he was a graduate student at CalTech and I was an Assistant Professor at UCLA. He was an 'honorary' member of my group, attending group meetings and classes, and was always in his lively, happy, and energetic mood. Everyone looked forward to his visits and provocative questions! He was an incredibly bright and kind person, and will be missed. Sincerest condolences to his family.
''Oh heart, if one should say to you that the soul perishes like the body,
answer that the flower withers, but the seed remains.'' K. Gibran
Tuesday, 19 January 2010
I learned so much from Sam. His mind was always racing with new ideas. He had high standards which I later inflicted on my PhD students. In papers and talks, every detail mattered; in research, he never settled for less than all-out effort.
Of course, Sam was not only a consummate researcher. He was also a remarkable person. He had a warm, broad smile and a constant sparkle in his eyes. He spoke alternately with passion and self-effacing humor. He could make a room of PhDs burst into laughter; he could also play gently with small children.
During our work on high dimensional data analysis, Sam and I experienced all the highs and lows that are part of any sustained research effort. In the spring of 2000, after months of flailing (which culminated in missing a critical conference deadline), we finally had our "Eureka!" moment. I have attached Sam's email on that day. I think Sam would approve: he loved to share the excitement of our field. I will always cherish the memories of our friendship.
Subject: *too* good
Date: May 17, 2000 4:55:40 PM PDT
This algorithm is absolutely the business. You must call me if you are still checking email.
*Everything* I tried it on worked. I am running out of data for this algorithm to chew up! I almost want it to fail, just to be sure it isn't some kind of trick...
1) The 1D manifolds, of course.
2) I ran it on the translated faces, and it nailed it.
3) I ran it on Josh's swissroll, boom!
4) I ran it on the 8x8 digits and it gave good results, in fact just as good as the best runs I ever had with anything else.
5) I ran it on the 16x16 digits and it gave great results, although hard to display, so now I am working on the geodesic code. That way I'll be able to interpolate between any two items.
6) I ran it on a database of many poses and expressions of a single person's face, again great. Waiting for interpolation on that too.
7) I ran it on a short movie, with the frames shuffled randomly. I told it to learn a 1d manifold. What do you think it did? Bingo, it sorted the frames into the correct movie order (although sometimes it sorts them backwards!)
8) It is running on text now.
Too bad this didn't happen a week ago. It is *perfect* for NIPS...
Sam was a great example to everyone who knew him, of someone who could be so professionally accomplished, and at the same time simply a modest, funny, really nice person. He will be missed.
Sam was a gem - as a thinker, as a colleague, and as a teacher. His first talk here, last September, was presented in a room overflowing with multi-disciplinary colleagues from departments throughout the university. Many of us sat on the floor, or stood out in the hallway, knowing that it would be well worth the discomfort. The talk was punctuated throughout with lively discussion, all ignited by Sam's brilliance, vibrancy and enthusiasm. Over the course of the Fall term, he came to quite a few of my group meetings, infusing each of them with that same boundless energy, penetrating and critical rationality, and the sheer joy with which he grappled with difficult problems. Even those who had never met him before were amazed and delighted by his spirit.
And from my brief encounter with the family side of his life, I have a clear image of him holding one of his daughters in our kitchen, showing her the clementines and various shiny utensils, and taking delight in her intent visual, tactile (and gustatory!) exploration.
I am grateful for the brief time I had with Sam. His death, so shockingly incongruous, has left me reeling. I hope we, his colleagues, collaborators, and friends, can find a way to understand it. Most of all, I hope that Meredith, and someday Aya and Orli, may draw strength and comfort from the affection we all felt for him.
Monday, 18 January 2010
I first met Sam in '99; I was a new Ph.D. student in Pittsburgh but spending my summers at the recently launched Gatsby unit. Sam, as has been said here repeatedly, was a postdoc there, where he stood out both academically and socially among a group that was exceptional in both dimensions (really!). Some years later, as a Gatsby postdoc myself, I often had occasion to think back on his example, and to try (and largely fail) to emulate it.
In his scientific work, I think Sam viewed himself as sort of a craftsman: collecting powerful tools, dissecting them, refining them, and applying them in new ways. He would speak modestly about this – once explaining that he had never written a proof, except insofar as he inherited the theoretical guarantees that were inherent in the tools themselves – but at the same time he believed strongly in this approach and was so clearly excited and driven by the possibilities of his toolchest.
That Sam’s work and his thinking were so grounded in this concrete, almost clockwork perspective was one reason that he was able to be so incredibly effective at communicating about them. Many of his memorable turns of phrase (which many of us have swiped) are also so physical: the goal of LLE was to build a “box” with “knobs” that would dial along the key dimensions of the input space; HMM algorithms could be described as involving “bees” flying between states, linear models as operating on “pancakes” in higher dimensional spaces. Although he made this, like everything, look effortless, Sam worked incredibly hard at developing the tools for communication much as he did for the research itself. Over the years, I remember watching him experiment with different presentational innovations. In the days of overhead projectors, he would construct sandwiches of multiple transparencies and slide them over one another to make animations. Once, he started his talk with the key to an elaborate color-coding scheme for his slides; my favorite was brown for statements that weren’t quite technically true: you could ask about the caveat afterward, if you wanted. One year, he discovered that if he laminated his NIPS poster, he could write on it with a dry erase marker. He took a great deal of pleasure, not just in the didactic possibilities of this, but also at the look of horror on the audience’s faces the first time they saw him whip out the marker and start scribbling all over his figures.
Much more important to me – though much harder to describe, and more painful to recollect – are Sam’s personal qualities. As others have mentioned, Sam served a sort of guide and provocateur for a group exploration of London’s nightlife (also the somewhat less adventurous nightlife of Whistler, BC, as others have wholly neglected). My first encounter with Sam’s legendary generosity was the way he would subsidize the graduate students' food and drink in our London outings – although he explained that this was a remnant of his socialist Canadian upbringing, I only later realized he was trivializing a fundamental trait of his character.
It was a pleasure and an honor to watch Sam grow up, always many steps ahead of me. Some steps were small. When he took his first faculty job, he decided that he really should no longer engage in his annual “NIPS flirtation” with a PhD student. Typical of his sometimes untethered spirit of sharing, he told me this by way of trying to set me (not yet faculty) up with the student in question.
Our final interactions all centered around the much bigger steps of family and fatherhood. Like so many of my colleagues, I was excited about all of the professional possibilities from Sam’s arrival at NYU. But as it happened, his arrival – to an apartment in our building – coincided closely with the birth of our son, so I saw little of him (or anyone) at work. Instead, Sam and Meredith were instantly and immensely helpful and thoughtful, repeatedly showing up unexpectedly at our door with various bits of advice and baby equipment we didn't yet realize we needed. Sam always patiently explained what each object was for, our son is now surrounded all day by Sam’s things.
Sam aspired to fatherhood before I realized it was something to aspire to, and it’s the calling that all his talents really point to. If I had to have a last experience with Sam, I am glad, then, (though “glad” is not really the word for somewhat less devastated) that it was this one. And if I have one regret that tops all the others, it is that I will never get the chance to learn more from Sam about how to be a father, as he perfects it all a few steps ahead of me; and of course that Aya and Orli, and Meredith, will not be able to enjoy his continued success at what, I am sure, would have been the thing he was most brilliant at of all.
Words simply can not express the immense respect that I have for Sam, who remains to this day the single most brilliant and caring person I have ever had the good fortune to meet. Perhaps my esteem for him can best be expressed by the fact that even though we hadn't seen each other in nearly 20 years, his tragic death has touched me and saddened me so profoundly.
"Shouks", my deepest condolences go out to you and to all of Sam's family. As a father myself, I can't even pretend to imagine how painful this must be for you. May you draw strength and comfort from knowing that all who knew Sam, even those of us who only knew him half a lifetime ago, shall be forever the better for having had that privilege.
The world has lost one of its very finest.
was so alive to you every time you met him. He had me off in whirl of ideas within seconds of meeting him 13 years
ago, and even the last thing I ever heard him say: "I feel like an email input-output node" seemed to capture something else perfectly. I'll miss that guy.
Sam touched me personally in a number of ways, a couple of examples which I'd like to share. When my son, Kiva, was considering attending UTS with some trepidation, Suzanne and I asked Sam if he could speak with him and share his own UTS experience. Sam's impact on Kiva was immediately profound, and witnessing Sam's passion and enthusiasm about the place changed his entire outlook on the prospect -- and hearing it from a guy that was infinitely cooler than his parents will ever be didn't hurt either. Kiva attended the school, and it was as wonderful as Sam said it would be. Although their paths would cross only occasionally over the next 6 years, Kiva was forever touched by Sam, and is feeling the same kind of hole the rest of us are. How precious are those that can affect us so deeply in such short a time. My other anecdote surrounded a decision I had to make as a committee chair on a particularly contentious issue. From his sabbatical in California, I received an email from Sam that meant so much to me. He sensed from afar how difficult the decision was for me, and through his incredibly kind and supportive words, lifted me from a state of feeling really discouraged to feeling like I'd done really well. I'll cherish his empathy, thoughtfulness, and sensitivity forever.
As I struggle forward with this, one thing is becoming very clear to me. And that is how incredibly fortunate I've been to have crossed paths with Sam. He's set an inspiring example for me of how to be a scientist, a supervisor, a teacher, a husband and father, and perhaps most of all, how to touch the lives of those around us. That inspiration and that example I'll carry with me forever.
Some recent memories of Sam include him and Meredith on the terrace of my house in Berkeley---it was a party for my research group but somehow it felt natural to invite Sam as well---and a long run on beach in San Diego---we ran for what seemed like forever to me and then I stopped and Sam kept running for a while longer.
I had a lot in common with Sam and I believe that he felt similarly. Both of us had a love for probability and for mathematics, especially the kind that could be explained in a few minutes to someone before their eyes start to glaze over. The kind of mathematics that was fun to discuss over a beer (well, a beer for me; Sam didn't need a beer to take his enthusiasm and joy up a notch). Both of us had a contrarian streak.
This similarity in outlook made me prize being able to run into Sam on a regular basis; I loved getting his take on new developments. When someone whom you basically agree with sees something a bit differently than you do, it's a great chance to learn something (perturbation theory, right Sam?).
Sam, my head is starting to spin again and I'm going to have to stop thinking about how much I'm going to miss you.
I continue to love being a part of the machine learning community, mainly because it brings together so people whose principal passion is a love for ideas, especially new ideas. They're friendly, but fierce, they're wise, but young. Sam was perhaps the paragon of this intellectual movement.
Sam, forever young.
His lectures were delightful; he was informal, enthusiastic, and always clear. He would show the audience something simple and
elegant, that in hindsight was clearly the right thing to do, although beforehand we had no idea.
A long review of a Neural Computation article that I co-authored ten years ago ended with "Good job, guys!". It reveals my image of Sam that I've always assumed those words were from him; it's the kind of thing he would say.
I've appreciated reading the other postings by his many friends. My heart goes out to Meredith and to all of Sam's family.
Sam, your scientific work showed us how local interactions could give rise to global implications. You gave us such joy as individuals, and also brought us closer together as a community. You will be forever missed, and we will not forget the young family you left behind.
Our reason for inviting Sam was 'merely' our knowledge of the important work that he had done. I for one had not actually met him, and I have to say that I was taken aback by the personal characteristics, namely dynamism, sociability, kindness and general 'nice-ness', that complemented his obvious creativity and enthusiasm for his work. He entered into the spirit of the activities in the Institute and tried hard to persuade us to incorporate the sort of 5-minute spots referred to by David MacKay!
I only encountered him for about a month and yet his departure seems to have left a massive gap.
What made Sam special to me was a shared passion for machine learning in audition, which was relatively rare. As I was developing new algorithms post MSR, I remember having in the back of my mind this thought: one day, soon, I'll meet up with Sam and tell him about them, and we'll have an awesome, totally unpredictable conversation. Then time suddenly leaped ahead... one day we'll be able to run a model on his lecture videos and research papers and play his part of that conversation. Sam, I wish you could be around when we do this problem, that would make it so much fun.
Sunday, 17 January 2010
what really strikes me most about him is his thoughtfulness, generosity and abundance of life. Sam inspires my faith in humanity when the world sometimes seems like a crazy place.
I first got to know Sam when we were both at Gatsby around 1999-2001. Afterwards Sam, Geoff, Max and I returned to Toronto more or less at the same time, so I was lucky to spend a year and a bit more with him.
I was fortunate to have worked with him on a number of papers and experienced Sam's enthusiasm and energy "behind the scenes". The way he motivated a grad student as myself and infused so much fun and excitement into the work was inspirational and one I aspire towards (never as well) ever since.
As many people have said, Sam has the most generous and selfless spirit. When I was applying for faculty jobs, the whole process seemed so incredibly daunting, so I asked various people for help and advice. Sam wrote me this amazing frequently-answered-questions list, like ten pages or something, detailing every step of the process, and all the little things you should watch out for, how should you pitch your talks, the politics within a department etc etc. It was awesome and really helped cool my worries and boost my self-confidence about being able to handle the whole thing.
I am little bit of a foodie so I have a few food related memories of Sam. Once we were having lunch at the tea room in Gatsby and Sam started applying peanut butter on his bread then slicing pieces of banana onto it to make a sandwich. I thought it was a little disgusting at the time but Sam loved it, saying how bananas have so much potassium and it's good for your brain. I tried it years later and it was great! Another time, we were at his place for brunch (this cool little apartment on the top floor of a converted house) and he made smoothies for people with this handheld blender.I thought "how cool is that!" and a handheld blender is indispensible in my kitchen ever since. Sam was the one who introduced us to chicken escalope at The Onion. I still go there once in a whilefor the chicken escalope on a brown bap. I'm happy to report that it's still as big as ever and enough for lunch and dinner!
Although these aren't really characteristic of him, they are fond little bits of memories I have of him, like the boom of his voice or the way his hands gesticulate, or the way he bounces towards the whiteboard in the midst of a discussion.
Sam, thank you for your warm friendship, I'll miss you.
Apart the usual NIPS parties, I remember this time with Sam when he was sitting in the back of our practical machine learning class taught at UC Berkeley in early 2008. Not surprisingly, he could easily blend with the other graduate students -- he was so accessible. We chatted a bit after the class about his current and past projects. He would preface them by some humble disclaimers that he "is just toying with things", nothing grandiose, but then would passionately talk about his ideas, with such energy that you couldn't stop yourself but join him in his enthusiasm. I was surprised at the time by his academic modesty, given how he was shining as an academic star. While talking about his future plans (would he stay in the Bay Area? would he could go back to Toronto?), his family would always seem to come up as the highest priority. We exchanged a couple of emails afterward about a possible visit I could make to Google. And in each exchange, I could feel his genuine human warmth. At the time, I actually had to make a big decision about my post-doc plans, and he selflessly offered that I call him if I needed advice about the different places that I was considering. I am actually quite thankful for the guidance he offered at the time.
I am still in shock with the tragic news. Sam was not just a brilliant researcher, but he was also an amazing person who would leave his positive mark in the people's lives who would be fortunate enough to cross him -- as the many posts below testify. I send my most sincere warm thoughts to his wife and two daughters -- there are no words to describe something like this. I would also like to echo Andrew McCallum's thoughts on the importance of the human dimension in our research community. Even away, Sam could still be building our community...
When he visited CMU a while back, Sam suggested having a joint CMU/Toronto machine learning get together, a great Sam-esque idea that unfortunately never happened. I worked with him to put together a proposal for discussion articles in the Journal of Machine Learning Research. Sam saw a bigger picture, a bigger community, than most of us. His intellect, charm, enthusiasm and energy contributed tremendously to making our field so vibrant.
The fact that I was not among Sam's many close friends, not a direct collaborator, probably helped make me a good choice for one of his letter writers for his move to NYU. That letter begins with the sentence "The preview is that Sam is one of the very best machine learning researchers active today, and certainly one of the stars of his generation; any department would be fortunate to have him." The middle of the letter discusses his work with Lawrence on LLE and remarks "If I were asked to identify just a few pieces of work as being among the most important in the field of machine learning in the last decade, this would certainly be on the list". The conclusion states "You should get Sam if you can."
Indeed NYU was fortunate to get Sam... and as I spend a lot of time in NYC, when I learned he was going there I was excited to think I might see him more often, professionally and socially. Like too many things in a busy life, I waited too long.
Goodbye Sam... it's clear from the posts here that you'll be remembered with limitless affection.
In 2000 Sam worked for a summer with us at WhizBang Labs in Pittsburgh. In the midst of the constant push and pressures of a start-up company, Sam was a breath of fresh air. Bubbling with ideas, he engaged with everyone from fellow researchers (like Fernando Pereira, Tom Mitchell, William Cohen, Drew Bagnell and Jonathan Baxter) to software engineers and sysadmins.
After I moved to UMass Sam visited several times. Three memories stand out.
Sam had just finished giving a delightful hour-long talk on Neighborhood Component Analysis to an audience of about 70 people. Many researchers might have been a bit tuckered out after making such a presentation, but Sam was still going strong, answering questions from a group of graduate students that had gathered around him afterward. When we realized that there was an hour-long hole in his visit schedule, he suggested that the ten of us stay in the room and continue talking together. We wandered over to a whiteboard, and the discussion ranged more broadly, eventually arriving at the relation between Bethe free energies and belief propagation. Sam proceeded to give an impromptu lecture on the subject---his gesticulating hands drawing pictures in the air, equations flowing out of his blue pen---all so clear, so beautifully motivated, unrolling as naturally as a fairy-tale. (One of the other blog writers had the perfect turn of phrase: "you couldn't help but understand" when Sam was explaining something to you.) Sam also pulled everyone into the conversation. In true Socratic fashion, each stage of the mini-lecture was driven by questions which Sam would ask the students (directed to the student by name, because Sam had taken the time to introduce himself to each student and learn their names). Somehow Sam's boyish enthusiasm prevented even the most junior students from feeling too self-conscious. At the end of the extra hour the students were filled with the sense of "belonging" that came with this new knowledge, and with Sam's infectious joy in it.
That weekend Sam stayed at my house. Unnecessarily, but totally characteristically for him, he brought gifts---a book for each family member: myself, my wife and two children. I can still see Sam sitting on the living room floor cross-legged, the small body of my four-year-old nestled in his lap, while Sam read aloud.
We had made plans to go sailing together on Saturday, but there was no wind. We decided just to spend a simple afternoon at home. I remember walking out on to the porch seeing Sam sitting there, feet up on the railing, eyes closed, face toward the sun. A quiet moment. Relaxed.
In 2007 I was asked to be program co-chair of ICML. I was going to beg off until I understood that Sam would be my co-chair partner. There is no one with whom I'd rather take on such a job. Sam and I became like climbing expedition partners---taking turns hauling the supplies, calling out advice about where to drive in the next piton, encouraging each other. He brought his creativity and boundless energy to this service task. I count literally over 1000 email messages from Sam during our 8 months of ICML work. Sam wasn't able to make it to Helsinki---his twins were born soon after ICML---but Sam's imprint is all over ICML 2008, and he initiated multiple innovations in the conference that are on track to persist for many years to come.
Sam, we miss you terribly. Know that you are loved dearly.
Meredith, I'm sorry we haven't met yet; I wish you peace, strength and courage. Reach out to Sam's friends, and we will be there for you.
For me the tragedy of Sam's death really brings to mind the value and importance of the personal connections we have in our research community---the enduring friendships, the caring, the human dimension beyond our research---and how much these friendships mean to me and all of us, even though we are scattered across the globe. Our grief over Sam's loss brings us together, and will deepen and strengthen our relationships. So, although we desperately wish Sam were here with us instead..., from the beyond, Sam is still building community.
Sam also initiated a tradition of entertaining the community, usually at holiday-time, with well-chosen puzzles, some of which I am still chewing over. Here's a message from Sam.
From: Sam Roweis
cc: "David J.C. MacKay"
Subject: coin puzzle
Here's a summer coin flipping puzzle to keep you occupied:
"You are given a biased coin, whose bias you do not know.
Using a sequence of independent flips of this coin, simulate
a sequence of independent flips of a fair (unbiased) coin."
The goal is to do this as efficiently as possible, in the
sense that you use as few flips (on average) of the biased
coin per simulated flip of the unbiased coin.
Here's an inefficient but correct way to get you thinking:
Flip the biased coin twice. If it comes up HH or TT, try again.
If it comes up HT, output H.
If it comes up TH, output T.
Of course if the bias is near 0 or 1 you will be rejecting a lot...
Sam, you were a wonderful, wonderful colleague and friend. Life was always filled with fun and humour and interest and intensity when you were around. We will miss you terribly, but you are not gone - you are still with us now, in the memes with which you have infected us.
Saturday, 16 January 2010
I met Sam in graduate school, when we were both in John Hopfield's group. It was an amazing time. We were all young, learning what it was to be a scientist. Sam delighted in the process, and delighted in reminding people of the joy of having so much to learn. In my mind, I can still hear him saying, "Ah, young butterfly, haven't you heard that the trace of the matrix is the sum of the eigenvalues? Let me show you why..."
He took such joy in seeing people understand - in seeing people happy, for whatever reason. He loved sharing intellectual jewels the way some people love telling jokes. We'd seldom see each other without his telling me about the latest mind-bogglingly beautiful result that he was enamored of. Maybe this week it's Propp & Wilson's exact sampling from markov chains, with applications to tilings, explained enchantingly in terms of a time machine and living backwards. And if I was too slow to catch the explanation in real time (I usually was) then it wouldn't be long before I'd get an email from Sam, spelling it all out in detail.
Sam loved sharing himself to make people happy. He was a people person. I had never heard of the phrase, "people person", before I met Sam - and if I had, I wouldn't have thought that a people person could also be a deep intellectual. Doesn't deep understanding entail privation and solitude and a retreat from the softer emotions? But knowing Sam disabused me of that conceit.
He was also perfectly willing to demand something of people - if the demand came with caring and love. Like the demand to take care of the world. Another concept Sam introduced me to was dumpster diving. But not dumpster diving as a cheap way to fill your fridge with perfectly good food the supermarkets were throwing away. Rather, it was clear he was talking about dumpster diving as social activism in protest of our society's wasteful ways. That, and the cup hanging by a carabiner from his enormous backpack.
I don't know if I've met anyone else who had the goodness of humanity so deeply embedded in his heart.
We'd have discussions about the future of society in the wee hours of the morning. Sam wasn't afraid to see that our world is changing, and that we should prepare for it - no: that we should take a role in guiding it for the better. I remember talking about the ever-increasing presence of surveillance cameras and email snooping and Big Brother and how it scared me, and Sam countering by explaining why he strived to live a life of openness with no secrets and no shame, because he saw the only solution being to prevent the government from being the ONLY ones with such pervasive knowledge; everyone must have access to the same information, so we must learn to live wisely in a world with no secrets and no privacy. That made me think, for a long time. I was surprised by his conclusion, and also surprised by his confidence that he could live a life so clean, so pure, that with no privacy would come no shame. I think he always strived for that kind of purity of feeling.
Many of my memories about Sam come from our time working together in the early days of DNA computing. When Paul Rothemund and I came back from the first conference on DNA-based Computers flushed with excitement, Sam joined us in the adventure and soon the three of us were meeting every week or so with Len Adleman and his friends at USC. We would have wide-ranging talks about life and science on our long drives to and from Len's lab or home. Sam was irrepressible, stripping our ideas down to their simplest and cleanest essence, suggesting new possibilities and new ways of looking at things, breaking our mental roadblocks with his insistence to understand: "Why?" and "Why not?" And always he would urge us on with his enthusiasm, or buoy us up with wise words when we were discouraged, overwhelmed, or depressed. (I sure could use a dose of Sam right now.)
To Sam, science was personal. So anyone could do it. With whatever was handy. Of such spirits was born the Sunday Experiment Club, with the charter to think about and do something unrelated to our thesis research every weekend. Alas, we only met twice - but the inspiration remains, as do memories of trying to move fluorescent DNA around in a puddle of saltwater by means of a 9V battery and two electrodes.
Of course, we all know that DNA computing wasn't Sam's destiny and it wasn't his true passion. I guess it was a hobby - a chance to drink more broadly of the scientific experience. And also sometimes I wonder if he did it just to be with his friends, us... maybe it was as much fun for him as stealing a city transit bus for a scavenger hunt? I don't know how he managed to squeeze it all in between writing "EM algorithms for PCA and SPCA" and "A unifying review of linear gaussian models". Sometimes I felt guilty that our collaboration took so much of his time away from his true passion. Perhaps more surprising is that at times it seemed that only his friends could see that machine learning was his true passion, and we had to convince him not to give up on it when he was afraid he would have nothing great to contribute.
One wants to remember Sam as a shooting star, blazing through the night sky with inspiring purpose and direction. But it wasn't always so. Like any mortal, at times Sam grappled with finding meaning and direction in his life. He seriously wanted to be an astronaut, for example, and was torn by the thought that being an expert in machine learning was not a ticket to the sky. What I admired so much about him, though, was how he never stopped thinking about what life meant, what principles to live for, how to do right by others - whether or not he knew any answers. He was a serious and sincere person in the most fun-loving way.
Sam loved to entertain us with stories, too. He had a lot of them, each delightful. Here's one, as best I remember it: Halloween. Sam's a young kid. Dad wants to make him a real special costume. 1 raincoat. 1 umbrella. 1 leaky "cloud" on a stick. Attach above umbrella. 1 basin for collecting water. Install around legs. 1 water pump, battery powered. Place in backpack, route tubing from basin to cloud. And little Sam walks from house to house with a rain cloud over his head! What fun!
P.S. For the geeks out there, I'll include one of Sam's signature files from grad school days.