essay on allen institute

Natural Language Processing

Computer vision, ai for the environment, experimentation and communication, our mission is to contribute to humanity through high-impact ai research and engineering..

AI2 is a non-profit research institute founded in 2014 with the mission of conducting high-impact AI research and engineering in service of the common good. AI2 is the creation of the late Paul G. Allen, philanthropist and Microsoft co-founder.

Headquartered in Seattle on the shores of Lake Union, AI2 employs the world’s best scientific and engineering talent in the field of AI, attracting individuals of varied interests and backgrounds from across the globe. AI2 prides itself on the diversity and collaboration of its team and takes a results-oriented approach to complex challenges in AI.

AI2 has undertaken several ambitious projects to drive fundamental advances in science, medicine, and conservation through AI. Meet our team , directors , and advisors , learn about our projects, and check out our current openings to see if there is a position that matches your interests and skillset.

Our Founder

Paul G. Allen launched the Allen Institute for AI in 2014 to explore critical questions in the field of AI. His philanthropy allows us to push the boundaries of fundamental AI research and to find high-impact ways to deploy these new technologies for the benefit of humanity.

Our Core Values

Our mission is to build AI for the common good. We aim to produce breakthrough research and tools that move the needle in AI, empower the research community, and benefit society.

Accountability

We take an efficient, results-oriented approach to our work. We define ambitious, timely goals and continually measure our success against them.

Transparency

We are clear and forthright with our team about our actions, our decisions, and the goals of our institute. We produce and share new research, tools, and resources openly with the wider world. We support open science.

Collaboration

We proactively develop diverse teams with a wide range of perspectives and skills to cultivate the best possible ideas. We build relationships with outside research groups and organizations to collectively elevate the field of AI and achieve the broadest possible impact.

Meet Some of the Team

"One of the best things about AI2 has been the variety of problems I get to work on to support AI2’s unique operational needs. On top of that, as a member of the queer community I feel incredibly thankful that I get to work in a place where my identity is respected by everyone without question."

"While there are many reasons to highlight why AI2 is a great place to work, undoubtedly the best part is the people who work here. Everybody's passion and dedication for AI2's mission is infectious. Oh–the free grapes are great too."

"There are bright, talented people in lots of organizations. Put them on a mission they care about, and you've got a potent cocktail for success and impactful contributions. That's why the vision and leadership at the heart of AI2 are key ingredients that make us thrive."

“I get to be surrounded by a creative, collaborative, interesting group of people, and I love it! This team know how to bring the heat - innovation-wise, and with their expert hot-sauce tolerance skills.”

"The one thing that I LOVE most about AI2 is my amazing colleagues. Everyone is smart and creative, yet very kind."

Natural Language Processing

Computer vision, ai for the environment, experimentation and communication, fine-grained human feedback gives better rewards for language model training, how far can camels go exploring the state of instruction tuning on open resources, realtime qa: what's the answer right now, swiftsage: a generative agent with fast and slow thinking for complex interactive tasks, scirepeval: a multi-format benchmark for scientific document representations, a question answering framework for decontextualizing user-facing snippets from scientific documents, crystal: introspective reasoners reinforced with self-feedback, demystifying prompts in language models via perplexity estimation, do all languages cost the same tokenization in the era of commercial language models, editing common sense in transformers.

My Chapter in AI  Storytelling at ICT

Dr. Melissa Roemmele is a Research Scientist on the Storytelling team at Midjourney, Her work explores the use of AI and NLP techniques to augment human creativity. Dr. Roemmele first came to ICT as a summer intern in 2010, and returned as a Graduate Research Assistant (2012 – 2018) while studying for her PhD under Professor Andrew Gordon, head of the Narrative Group. In this essay to celebrate ICT’s 25th Anniversary, Dr. Roemmele traces the throughline of her research from her time at ICT, to her work today.  

In 2010, the possibility of AI systems telling stories still seemed very far away. I’d just completed the first year of my MA program in computational linguistics and came to ICT as an intern, hoping that it would prepare me for a job developing software for natural language processing. I had no idea that it would profoundly shape my entire career trajectory.

AI storytelling was the vision my internship advisor Andrew Gordon conveyed to me as the foundation for his research and my prospective internship project. Today, powered by large language models (LLMs), apps like ChatGPT can generate stories that have many of the same qualities as human-authored stories. However, Andrew and his collaborators started working toward this goal decades ago, long before most people were even aware of the idea.

Andrew identified that progress was hindered by a lack of evaluation benchmarks through which progress could be demonstrated. My internship project addressed this gap by creating a framework to quantitatively measure comprehension of commonsense relations between events in stories. This measure, called the Choice of Plausible Alternatives (COPA) , was one of the first benchmarks for assessing story understanding via natural language. The COPA framework  continues to be widely used to evaluate LLMs today.

Using GenAI to Tell Stories

Recognizing the unique research opportunity at ICT, I returned in 2012 as a PhD student in Andrew’s lab. This coincided with breakthroughs in neural network training, which revolutionized language modeling and its applications in various NLP tasks. In contemplating my dissertation topic, I identified the opportunity to examine how language models could be used for AI-based storytelling.

As part of my dissertation research, Andrew and I built the application Creative Help, which was an early demo of using generative AI to help people write stories. Creative Help used a language model trained on web-sourced fiction to suggest new sentences for authors to incorporate into their stories. This language model was fairly primitive compared with today’s LLMs, so the authors who used Creative Help didn’t consistently find the content of the suggestions helpful. However, many of them became excited about the future prospect of AI-augmented writing as a result of trying it. Of course, this prospect has since become a reality.

The Creative Help interface allowed us to analyze how the authors incorporated the AI-generated suggestions in their stories. Based on this data, we were able to identify some basic linguistic features that made suggestions more appealing to authors. Still, this analysis only scratched the surface in capturing people’s complex perception of the AI-generated content. I had learned the critical role of evaluation in AI research from my internship project on COPA several years earlier, but my work with Creative Help further showed me that evaluating AI systems can be just as challenging as building them. For one thing, human users bring widely varying objectives to their interactions with systems. For example, some Creative Help authors were disappointed that the system’s suggestions lacked coherence with their story, whereas others enjoyed this incoherence because it steered their story in a surprising new direction. I came to appreciate the difficulty of designing evaluation methodologies that try to measure quality of AI output according to objective criteria while still accounting for diversity in subjective dimensions of judgment.

Applying GenAI Research 

After finishing my dissertation at ICT, I joined the Language Weaver division at SDL, now part of RWS. As part of a product-oriented team, I had the opportunity to work on language generation tasks adjacent to AI-based storytelling, including summarization and translation. From this experience I gained a stronger perspective on how to leverage NLP to build software to solve specific user problems.

Much of my work at RWS involved developing methodologies for readily evaluating the output of generation models. It was particularly challenging to establish quality metrics for models where no such capability previously existed, and additionally to resolve conflicts in quality signals. This evoked the same difficulty I encountered in assessing the highly ambivalent user interactions in Creative Help. I realized that my research at ICT helped me more readily see uncertainty in data not simply as a barrier to modeling and interpreting the data, but as an opportunity to propose and explore new research questions.

GenAI Augmenting Human Creativity

In June 2024, I joined Midjourney as a research scientist on the Storytelling research team, where we explore the use of generative AI tools in augmenting human creativity. I’m thrilled to once again be immersed in this research space, now that there’s been so much recent progress in expanding core AI capabilities. With this progress, there are new opportunities to examine fundamental research questions about how to optimize human-computer interaction.

It’s clear the path to my current role began with that ICT internship nearly fifteen years ago. Reflecting back on that time, when AI-based storytelling was still mostly hypothetical, it wasn’t always clear how the work I was doing would help make it a reality. Much like a well-composed story where early events gain significance through later developments, I can now make much more sense of how my efforts at ICT have contributed to a larger research vision. And I feel very fortunate that I haven’t reached the end of this story yet.

ALLEN TALLENTEX 2025: ₹150 Registration Fee Offer Ends 30 June 2024

TALLENTEX 2025: Your Gateway to Success

👉 Cash Prizes for Over 15,000 Students 👉 Cash prizes worth ₹2.50 crores and scholarships worth ₹250 crores 👉 Exam Accessible Offline, Online, and via mobile

The country’s renowned talent encouragement exam, TALLENTEX 2025 , has been announced. A large number of students register for the exam every day. In response to the students’ excitement for the exam,

ALLEN is offering a 50% discount on registration until 30 June . The 11th edition of the exam will be held both offline and online, and students can also take it via mobile.

TALLENTEX provides a platform for students to showcase their talents and helps them on their journey to achieving their academic goals.

Students can also receive scholarships for ALLEN Classroom and Online Courses by participating in this exam.

Since its inception, 14.75 Lac students have participated in the TALLENTEX exam . Students from classes 5 to 10 can appear online and offline in this exam.

The online exam will be held between 5 to 20 October across various slots, while the offline exam will be conducted on 13 and 20 October.

In addition to national and state rankings, cash prize of ₹2.50 Crores and scholarship up to ₹250 Crores will be awarded. 4,750 cash prizes are for offline, and 10,500 cash prizes are for online.

Students can also receive up to a 90% scholarship for admission to ALLEN classroom and online courses. A separate Competitive Success Index will be issued to every student.

Students can register online at www.tallentex.com . Offline forms are available at ALLEN centers. The results and prize distribution will be announced in November as part of the Success Power Session.

Exam Format

The exam will last two hours and will include multiple-choice and integer-type questions covering Physics, Chemistry, Math, Biology, and Logical Mental Ability.

The paper is based on the NCERT syllabus. Through TALLENTEX , students can get assistance in choosing their subjects and self-assess their preparedness for various national-level exams.

A Better Platform for Self-Assessment

Students participating in TALLENTEX will receive a Competitive Success Index, allowing them to gauge their potential national ranking in exams like JEE, NEET, CA, and CS.

This index will be issued with the result announcement. The exam helps students analyze their academic excellence and measure their skills through healthy competition with thousands of peers.

Scholarships and cash prizes based on talent are also awarded, aiding in career development.

Free Practice Papers

Upon registering for TALLENTEX , students will receive free test papers for practice available on the website, helping them understand the test level and prepare accordingly.

The exam syllabus follows NCERT, and information about the exam pattern and syllabus is also available on the website .

Recommended Read:-

• TALLENTEX 2023: Examination Held in 11 States
• Cracking The TALLENTEX Exam: A Detailed Study Plan
• The Role of TALLENTEX in Boosting Your Child’s Career
• Revealing the Rewards of TALLENTEX 2024: More Than Just Cash Prizes
• Understanding TALLENTEX: A Complete Guide For Students and Parents
• TALLENTEX 2024: 3,16,000 Students Participated in the 10th Edition of Exam
• ALLEN TALLENTEX 2023: Over 1.48 Lakh Students Appear for the Second Round
• TALLENTEX 2024 Announced – Students Will Get Scholarship Worth Rs 250 Crores & Rs 1.25 Crore Cash Prize

Related Posts

 Cash Prize to More than 15 Thousand Students Excitement is in the air as ALLEN Career Institute Private Limited has…

• Search Please fill out this field.
• Manage Your Subscription
• Give a Gift Subscription
• Newsletters
• Sweepstakes

• Celebrity Family

All About Jeremy Allen White's Parents, Richard White and Eloise Zeigler

Jeremy Allen White credits parents Richard White and Eloise Zeigler with his passion for acting

Jessica Sager is a contributing writer at PEOPLE. She has been working at PEOPLE since 2023. Her work has previously appeared in Parade, The New York Post, Seventeen, PopCrush and more.

Kyle Goldberg/BFA.com/Shutterstock 

Jeremy Allen White adores his parents, and his mother has played key role in his career.

White stars as Carmen "Carmy" Berzatto in The Bear , and his mother, Eloise Zeigler White , keeps him abreast of fan thirst for him and the character.

While he eschews most social media (save for the occasional Instagram post), he recently told Vanity Fair , "My mom tells me what they're saying on Twitter, which is nice."

White's father, Richard White, and Zeigler live in New York City and keep a low profile, but the Shameless alum's mother has been in the spotlight with him occasionally. Zeigler has made several appearances with the actor, including on coffee outings and even on the red carpet at the 2023 Screen Actors Guild (SAG) Awards.

When White won his first Emmy award for his performance in The Bear in January 2024, he included his mom and dad in his acceptance speech.

"My parents. I love you, thank you for your support," he said.

Get to know Jeremy Allen White's parents, Eloise Zeigler and Richard White.

They were fledgling actors

Kevin Mazur/Getty

White's parents both used to be actors, and they met when his father attended one of his mother's plays in New York City's Upper West Side, White told GQ U.K . in October 2023. While they pursued Broadway careers, once White was born, his father took a day job recording depositions.

"Acting was something I was always aware of, it was something both my parents did, when they were in their 20s and 30s," White told Fault in 2022. "They met in [New York City] doing theater, there was always a passion for theater, film and the art of acting in our household growing up."

White followed in their footsteps and did theater when he was younger, and recalled to Variety the differences between his stage and screen experiences.

"The repetition of theater and doing the same thing over and over in front of people — you build confidence that way," he said. "I think for television and film, the bummer about it is you sit around and then you have five minutes to make something happen. I leave set every day just uncertain if we did what we were supposed to do."

His family was based in Brooklyn

Jeremy Allen White Instagram

White's parents raised him and his younger sister in Brooklyn's Carroll Gardens neighborhood, and White still keeps a home base in the New York City borough.

"I grew up in Brooklyn, we have family here. We try to come for the summers to see friends and family while here," he told Fault in 2022. "I like those New York nights where the sun can go down but it’s still very hot. Maybe I’ve over romanticized it a little bit, and it’s really miserable. But I like it here during the summer."

His mom is originally from North Carolina

Zeigler is originally from North Carolina, and her background helped White connect with another actor he admires, Dexter star Michael C. Hall , who he ran into at a holiday party.

"I was really nervous, and I happened to know too much about him," White told The Chicago Tribune . "I was like, 'My Mom is from North Carolina. You're from North Carolina. My Mom went to [the same school as you].' Once I got it out of my system, I calmed down a bit. He was very complimentary of [ Shameless ]."

His parents encouraged him to participate in physical activities

White says that when he was a child, he struggled to focus in school, leading his parents to "throw me into anything physical."

White played baseball, lacrosse, soccer and wrestled, but his favorite activity was dancing.

"I liked the performance of it," he explained to Fault. "I found that I was able to focus while performing, which is something that I found difficult to do especially as a child, even still to this day."

His mom was his date to the SAG Awards

Amy Sussman/WireImage

Zeigler accompanied White to the Screen Actors Guild (SAG) Awards in February 2023, and White told GQ U.K. that on the way to the red carpet, she nervously fan-girled over the idea of meeting Cate Blanchett . White recalled he assured Zeigler that she wouldn't meet Blanchett, but they did have a run-in with the Australian Tár star, who noticed that Zeigler was "crying real tears."

White said that Blanchett consoled Zeigler and told her she was a fan of White.

"I'm like, 'You must have better things to do [than watch me],' " White said he told Blanchett. For her part, Blanchett insisted the opposite was true.

"One can tell so much by meeting an actor's family, and to meet Jeremy and his mother simultaneously was a godsend," Blanchett told the magazine over email, adding that she is now a producer on Fingernails, starring White. "I can only thank his mother Eloise for creating this opportunity."

His father inspired some of his jewelry

Frazer Harrison/WireImage

If you look closely at White's hands, you may see him wearing a ring engraved with a W. In a sitdown with GQ , he said that the ring was the first piece of jewelry he ever bought, and he initially didn't plan on keeping it for himself.

"I actually bought this for my father for Christmas, but he's got really big fingers and they didn't fit," he said. "So it became mine."

Another piece White wears often is a subway token necklace, which he said was inspired by growing up in New York City and his father always having a jar of the tokens in their home.

"Now everybody in the family has one," he said, naming his mother, sister, father and estranged wife as all having one of their own.

He says they aren't great cooks

Matt Dinerstein / FX on Hulu / Courtesy Everett Collection

While White may have taken after his parents in terms of his love for performing, his cooking skills aren't a family trait. White attended culinary school to prepare to star as chef Carmy in The Bear , and he told Falstaff that growing up, his parents weren't exactly Michelin material.

"My parents are lovely, but they are not good in the kitchen," he said. "For me, the ultimate comfort was grilled cheese and tinned tomato soup. When I was sick, that was always a very comforting meal."

These days, White says that he does most of the cooking for family holidays.

"I did Christmas Eve dinner, I did Christmas Day dinner for me and my family — all big meals," he told PEOPLE in August 2022. "I love a Sunday roast. I love something really hearty, almost wintry comfort food. I'll do short ribs, roasted vegetables, some kind of potato. That's what I like to make and what I like to eat."

Related Articles

Allen School News

For these nationally recognized allen school undergraduates, research impact is its own reward.

Earlier this year, the Computing Research Association honored a select group of undergraduate students from around the country who have made notable contributions to the field through research. The CRA Outstanding Undergraduate Researcher Awards competition historically has been good to Allen School students. To the four most recent honorees — award winner Kianna Bolante , finalists Claris Winston and Andre Ye , and honorable mention recipient Nuria Alina Chandra — even more rewarding than national recognition is realizing the impact their contributions can have on individuals and communities in Washington and beyond.

Kianna Bolante: Empowering students to think critically about technology

As a member of multiple groups underrepresented in computing, CRA award winner Kianna Bolante is always on the lookout for ways to blend her curiosity about technology with a desire to create positive social change. 

“This mindset has guided me towards opportunities aligning with my values,” she said.

In addition to multiple service and leadership roles on campus, those opportunities have included research projects spanning accessibility, robotics and social computing. Bolante’s interest in the latter led her last year to pursue what is arguably her most high-profile work to date: a comprehensive social computing curriculum aimed at guiding middle and high school students to consider the different ways they and others interact with technology. Her interest in the project was inspired in part by her realization that, although students spend increasing amounts of their time online, most secondary education curricula do not prepare them to engage meaningfully on the topic of how people interact with digital technologies.

“A community’s core comes from how its individuals interact with each other, and social computing systems are a vital network affecting how humans build community daily — even more so due to the COVID-19 pandemic,” observed Bolante. “Yet, despite its pervasive influence, social computing topics are not explicitly taught in secondary-level courses in the U.S.”

Keen to bridge that gap, Bolante teamed up with professor Amy Zhang and members of the Allen School’s Social Futures Lab to develop six educational modules that teachers could incorporate, in whole or in part, into their classroom lesson plans. The modules, which covered topics spanning online behavior, machine learning and bias, misinformation and more, offered a combination of lecture content, hands-on activities and resources to support in-class discussion as well as personal reflection. After creating the curriculum, Bolante proceeded to travel around to Seattle-area schools, where she shared the lesson content with more than 1,400 students and their teachers. 

“That she has impacted so many young people is not only a show of her strong ability to lead and communicate but also the dedication she brings to educating young people,” Zhang said. “Kianna took modest expectations for the project and blew them out of the water!”

That dedication would lead Bolante to lead-author a paper reporting on the results of pre- and post-lesson surveys . She presented at the Association for Computing Machinery’s Technical Symposium on Computer Science Education ( SIGCSE ‘24 ) in March.

“This area of study affects us all through our constant use of collaborative systems and social technologies, such as social networking platforms, but we don’t all experience them all the same way,” Bolante noted. “We incorporated culturally responsive pedagogy in our design to help students connect their learning to their identities. A constant theme is to encourage students to think critically on the positives and negatives of social technology designs.”

Zhang and her Allen School colleague Maya Cakmak were delighted, but not surprised, that the CRA chose to recognize Bolante for her contributions with an Outstanding Undergraduate Researcher Award. Bolante had enrolled in Cakmak’s introductory research seminar as a sophomore, following an initial foray into research via a project analyzing language preferences in relation to disability and her work with Zhang through the DUB REU program during her first year at UW. Although the seminar is designed for those with no previous research experience — and Bolante already had a published paper to her name, after the aforementioned analysis appeared at the International ACM SIGACCESS Conference on Computers and Accessibility ( ASSETS ‘22 ) — she hoped to hone her independent research skills and connect with other like-minded students. 

Bolante subsequently returned the favor by serving as a teaching assistant in the next offering of the course, sharing her knowledge and experience. In addition, she has participated in extensive outreach activities as a member of various student groups on campus in addition to her continued engagement with middle and high school students. She is also blending her interests in education pedagogy, social computing and accessibility in her work with another Allen School professor, Kevin Lin , to incorporate content that reflects a functional understanding of different physical, sensory and cognitive abilities into the school’s own Data Structures & Algorithms course. Her many contributions inspired her inclusion in the 2024 class of the Husky 100 , a designation that recognizes students who are making the most of their time at UW.

“Kianna has contributed to so many research projects. She is a quick learner and resourceful in solving problems and adapting to the needs of each project. She is also a clear communicator who truly cares about her audience,” said Cakmak. “It is no wonder she has given so many talks and is asked to be on so many panels — and she never passes on an opportunity to use her voice to empower others.”

Nuria Alina Chandra: Decoding the genetic basis of disease

Ever since she first set foot on campus, Nuria Alina Chandra wanted to try her hand at research into the mechanisms of disease. But the difficulty she initially encountered trying to break into a lab — figuratively speaking, of course — galvanized her interest in giving other like-minded students a leg up.

“I originally struggled to find research opportunities in college, so once I found a position, I wanted to help others to do the same,” said Chandra. “As an Undergraduate Research Leader, I connect first-year students from underrepresented backgrounds with research opportunities and advise them on how and why to launch a research career.”

Chandra’s own research career has followed some twists and turns on the way to earning a CRA honorable mention for her contributions. As it turned out, her willingness to take an indirect route appealed to professor Sara Mostafavi in the Allen School’s Computational Biology group.

“Nuria approached me to ask about possible research internships,” recalled Mostafavi. “I was immediately impressed by her strong academic foundation as well as her diverse research experience.”

That diverse experience predated her arrival at the Allen School, when Chandra spent nearly two years working in the Bacteriophage Lab at The Evergreen State College. She then spent a summer as a research assistant in the Subramanian Lab at the Institute for Systems Biology before joining Seattle Children’s Pediatric Sleep & Pain Innovations Lab , where she investigated the development of posttraumatic and postsurgical pain.

In Mostafavi’s lab here at the UW, Chandra uses machine learning to analyze the role of regulatory DNA in cell-type specific gene expression to understand the downstream effects of DNA mutations associated with hereditary diseases such as type 1 diabetes and multiple sclerosis. 

“Regulatory regions of the genome are difficult to study experimentally due to their cell-type-specific effects,” Chandra explained. “The ability to accurately predict the transcriptional consequences of regulatory variants is critical to decoding the genetic basis of disease and facilitating the development of personalized treatment systems.”

Previous research suggested that information about the distribution of chromatin accessibility, which regulates gene expression, at the level of DNA base pairs could offer useful insights into regulatory protein behaviors. Such insights would ostensibly improve deep learning models’ predictions about chromatin accessibility across cell types. When Chandra set out to test this hypothesis, however, she found that existing deep learning models like BPNet could only accurately predict the accessibility profiles of a subset of regulatory DNA; they weren’t up to the task of predicting differential accessibility for roughly 90 closely related immune cell types. So Chandra developed a new, more robust convolutional model, bpAI-TAC, that is capable of harvesting data from base-pair resolution accessibility profiles to improve regional accessibility predictions. 

After demonstrating that bpAI-TAC outperforms the previous state of the art on regional chromatin accessibility prediction, Chandra has begun to explore techniques for extracting additional biological insights into the mechanisms of disease. In the future, researchers will be able to use Chandra’s model to catalog the effects of genomic variants in order to better understand how they contribute to disease and to assist with targeted drug development.

“I typically don’t assign such a challenging project to an undergraduate, but given Nuria’s enthusiasm and performance on her course work, I took a chance,” Mostafavi said. “And she exceeded my expectations! I’ve been very impressed with Nuria’s dedication, analytical abilities and approach to solving complex problems.”

Claris Winston: Creating technology that removes societal barriers

CRA finalist Claris Winston’s interest in research is not just academic; it’s personal. Based on her own experience, Winston began to appreciate firsthand the extent to which societal barriers and inaccessible environments challenge people with disabilities. The realization motivated her to develop a digital assistant called MyScoliCare that has helped patients and therapists around the world manage their treatment of the condition.

“This experience emboldened me to develop impactful, accessible technology during my time as an undergrad,” Winston said. “As my awareness has grown, so has my passion to create technology that improves accessibility for all and removes societal barriers for people with a variety of needs.”

Before pursuing that passion, she got her feet wet doing research in a wet lab — specifically the Molecular Information Systems Lab , where she was lead author of a paper describing combinatorial polymerase chain reaction , a novel technique for selective retrieval of DNA oligo pools to enable DNA-based data storage at scale. With a published paper to her credit, Winston then teamed up with siblings Cailin, Caleb, Chloe and Cleah on a project that applied software engineering techniques to detect and repair faults in brain-computer interfaces. BCIs are implantable devices that augment or restore sensorimotor function in people with neurological disease or spinal cord injury.

“Claris knows how to seek out interesting research opportunities and is driven to engage in projects that are meaningful,” Allen School professor Jennifer Mankoff said.

One might even say Winston has a magic touch when it comes to computing research. In the fall of 2022, she joined Mankoff’s Make4All Group to build upon a pilot project that incorporated embroidery textures into scalable vector graphics. The goal was to develop a pipeline for producing embroidered tactile graphics that make visual media accessible to people who are blind or visually impaired (BVI).

“Common formats for tactile graphics degrade over time,” Winston explained. “Since they are made of fabric, embroidered graphics are more durable, can be washed, can be sent to low-resource regions, and may be used by different people over the years.”

Winston spearheaded multiple aspects of the pipeline, from ensuring the optimization algorithm would produce sufficient contrast between textures, to extending its capabilities to incorporate lines and points, to the actual printing and post-processing of the embroidered graphics. The latter required her to troubleshoot the process for printing on satin fabric, which is prone to wrinkles and slippage, in addition to developing a reliable approach for stitching on both sides to make it braille readable. Winston also took the lead in recruiting, running and analyzing data from user studies with BVI individuals. 

Her ability to drive the project toward a successful outcome, including a paper submitted to the journal Transactions on Accessible Computing (TACCESS), was even more remarkable considering the circumstances. Shortly after Winston joined the lab, Mankoff was compelled to take time off for a family obligation. Her mentor’s unexpected absence didn’t faze Winston, who excelled despite minimal supervision. She also happily extended her involvement beyond the expected end date she had originally been given to see the project through, culminating in the TACCESS paper submission, and is currently working towards a submission that, if accepted, would be her fourth published paper in as many years. Winston is also collaborating with other researchers to build a value based healthcare system that uses large language models. Her work on that project was accepted to a workshop at the IEEE International Conference on Healthcare Informatics ( ICHI 2024 ).

In the past, Winston has worked on research at the intersection of computing and biology. As a research assistant in the Matsen Group at the Fred Hutchinson Cancer Research Center, she contributed to the development of methods for efficiently indexing COVID-19 phylogenetic trees using annotated directed acyclic graphs for the purposes of tracking variants of the virus. The project is yet another example of Winston’s maturity and commitment to real-world impact — an impact that earned her an Outstanding Senior Award at the Allen School’s recent graduation celebration.

“Claris already functions more like a graduate student than an undergraduate and is destined to grow into a strong independent researcher,” Mankoff said. “And with her commitment to service, she is also someone who will contribute to the ongoing work of making the field of computer science more inclusive.”

Andre Ye: Getting philosophical to build better models

CRA finalist Andre Ye is philosophical in his approach to machine learning research. This can be taken literally as well as figuratively; by pursuing degrees in both computer science and philosophy, he aims to explore the intersection of machine learning and topics such as human subjectivity and critical thinking.

“I believe building stronger connections between these two areas is essential to building more robust and usable models,” Ye said.

His first project uniting the two sought to account for human uncertainty in the annotation of medical images for training computer vision models intended to assist with clinical decision-making — a high-stakes task that affects the quality of a model’s downstream predictions and, potentially, patient care.

But first, he had to convince Allen School professor Amy X. Zhang to allow him to pursue the project in her Social Futures Lab . It helped that he had read a recent paper from the lab suggesting a new approach for capturing calibrated uncertainty, which dovetailed nicely with his proposal inspired by a previous stint segmenting kidney tissue images at UW Medicine.

“Normally, I might not give such a young student so much free rein to start with, but Andre quickly showed his ability to independently make progress on research,” said Zhang. “And since neither I nor my Ph.D. student who helped mentor him had any prior experience with medical data, we really leaned on and learned from Andre, given his past work in medical imaging.”

Ye took the reins and ran with them. While conventional approaches attempt to compensate for human subjectivity by incorporating an uncertainty distribution drawn from existing samples, the results are difficult to interpret and don’t necessarily indicate clinical significance. As an alternative, Ye spearheaded the development of Confidence Contours , a novel framework for developing annotations that explicitly account for human-provided uncertainty across a range of possible segmentations that can be used to train any general-purpose segmentation model.

“Rather than providing a singular segmentation, annotators annotate both a region of high confidence and additional areas of lower confidence,” Ye explained. “By leveraging human subjectivity instead of working around it, our approach produces models that are more useful to clinicians than those that rely on standard annotations.” 

His efforts were rewarded with an honorable mention at last year’s Conference on Human Computation and Crowdsourcing ( HCOMP ‘23 ). Ye subsequently embarked on another project advised by Zhang, along with Allen School professor Ranjay Krishna , examining how variations in visual perception as a product of different linguistic and cultural backgrounds influence the output of image captioning models. In a paper submitted to the International Conference on Learning Representations ( ICLR 2024 ), the team described how multilingual annotations convey more varied and comprehensive information about the objects, their relations and attributes depicted in an image compared to monolingual ones. The analysis showed how even supposedly objective “ground truth” is, in reality, shaped by human subjectivity.

Ye’s latest work explores whether language models, which can assist people with accelerating or even automating rote cognitive tasks, can also be effective tools for deeper thinking. Based on interviews with philosophers at academic institutions across the United States, Ye conceived of the selfhood-initiative model, a framework for defining critical thinking tools that can serve as a basis for designing language models that can be used to help human ideas take shape.

He intends to continue his investigation into the intersection between machine learning and philosophy as a graduate student; in the opinion of his mentors, Zhang and Krishna, he is already operating at that level.

“Andre is an astonishing young scholar in every sense,” said Krishna. “It would be hard to find another student who matches his level of curiosity, creativity and ambition.”

In addition to Ye and his peers, another undergraduate researcher with an Allen School connection, Thanh Dang , earned accolades from the CRA this year. Danh, who is studying computer science and mathematics at Colgate University, received an honorable mention in part for her work with Allen School professor Michael Ernst on a summer project in which she developed an evaluation infrastructure for comparing the performance of commit untangling tools. 

For Allen School majors who are interested in pursuing research as part of their Husky experience, CSE 390R Introduction to Research in Computer Science & Engineering offers the opportunity to gain hands-on experience with typical research responsibilities before seeking positions with labs or external research organizations. The next offering of the course, in autumn 2024, will be led by professor Leilani Battle .

Visit the CRA website to learn more about the Outstanding Undergraduate Researcher Awards program.

At least eight die in inferno near Moscow, TASS says

• Medium Text

Sign up here.

Reporting by Gleb Stolyarov; Writing by Guy Faulconbridge; Editing by Ros Russell and Angus MacSwan

Our Standards: The Thomson Reuters Trust Principles. New Tab , opens new tab

World Chevron

Beryl sets precedent for 'very dangerous hurricane season'

Hurricane Beryl, which has strengthened to a Category 5 storm, is setting the tone for a "very dangerous" hurricane season, the World Meteorological Organization (WMO) said on Tuesday.

Albania's world-renowned novelist Ismail Kadare dies at 88

Renowned Albanian novelist Ismail Kadare has died after being rushed to a hospital in the country’s capital, Tirana

TIRANA, Albania — Albanian novelist and poet Ismail Kadare, whose irreverent works from inside communist Albania earned him international renown and repression from the country’s dictatorship, has died in Tirana, his publishing editor said Monday. He was 88.

Kadare won a number of international awards , and had long been mentioned as a possible contender for the Nobel Prize in literature.

Albanian President Bajram Begaj praised Kadare as the country’s “spiritual emancipator.”

“Albania and Albanians lost their genius of letters ... the Balkans (lost) the poet of its myths, Europe and the world (lost) one of the most renowned representatives of modern literature,” Begaj said in a statement released by his office.

Albania’s government declared two days of national mourning on Tuesday and Wednesday, when flags will fly at half-staff. A minute’s silence will be observed nationwide Wednesday following Kadare’s funeral.

Onufri Publishing House editor Bujar Hudhri said the author died Monday morning after being rushed to a hospital.

A nurse at the hospital said he was taken to the emergency room after suffering cardiac arrest. She spoke on condition of anonymity because she was not authorized to talk to the media.

Kadare became internationally recognized after his novel “The General of the Dead Army” — which later inspired a film starring Marcello Mastroianni and Anouk Aimee — was published in 1963. The book told the story of an Italian general who was sent to Albania to find and repatriate the bones of thousands of his compatriots killed there during World War II, and who dwells on the futility of the task and of war.

At the time, Albania was still governed by the communist government of late dictator Enver Hoxha that had turned the small, mountainous Balkan country into Europe’s most isolated.

Celebrated for the delicate writing of his novels, Kadare fled to France in the fall of 1990, just a few months before the collapse of the communist regime following student protests in December. He lived in Paris and had recently returned to Tirana, the Albanian capital.

During a visit to Albania last year, French President Emmanuel Macron awarded him the Grand Officer of the Legion of Honor title . France had previously also made him a foreign associate of the Academy of Moral and Political Sciences, as well as Commander of the Legion of Honor.

Kadare was awarded a number of international prizes, including the inaugural International Booker Prize in 2005. His works, which included more than 80 novels, plays, screenplays, poetry, essays and story collections, were translated into 45 languages.

Born on Jan. 28, 1936, in the southern Albanian city of Gjirokaster, Kadare graduated from Tirana University’s History and Philology Faculty and went on to study at Moscow’s Maxim Gorky Literature Institute.

But he was recalled after Hoxha split with the Soviet Union — the first of two great ruptures with major communist powers that was later to conclude with China.

Back in Albania, Kadare won a reputation as a poet and novelist, but soon fell foul of the communist regime, which banned several of his works and briefly exiled him to the provinces.

“The General of the Dead Army” attracted major international attention when it was translated into French and published in the West. This recognition abroad has been credited with shielding Kadare from the more violent retribution Albania’s communists routinely reserved for dissidents.

After the fall of communism in Albania, Kadare resisted calls from different political parties or politicians to become the country’s president.

He is survived by his wife, Helena, also a writer, and his daughters Gresa and Besiana.

The funeral will be held on Wednesday.

Semini reported from Bari, Italy.

• Corpus ID: 128849088

COLOR TV BRIDGE: PARIS TO MOLNIYA-1 TO MOSCOW,

• A. Fortushenko
• Published 15 December 1967
• Engineering

Related Papers

Showing 1 through 3 of 0 Related Papers

Join our team of passionate explorers.

Help us advance science with the potential to change lives.

We are Allenites:

neuroscientists, cell biologists, geneticists, immunologists, computer and data scientists, informatics experts, mathematicians, engineers, and science support professionals

Our strength comes from teams of individual experts all working together toward our common goal: improving human health through the power of science. The Allen Institute is based in Seattle’s South Lake Union neighborhood, the city’s hub for technology and biomedical research. Our 270,000-square-foot headquarters and state-of-the-art research complex facilitate the collaboration and cross-disciplinary interaction that fuels our science.

VIEW JOB OPENINGS 

Intern & Postbaccalaureate Programs

The Shanahan Foundation Fellowship at the Interface of Data

Success starts with inclusion

At the Allen Institute, we believe that science is for everyone – and should be open to anyone who’s passionate about discovery and improving human health. We recognize that science has a long history of racism and exclusion, and we are dedicated to combating those biases within our own workforce and helping to reduce barriers to STEM careers more broadly.

We also believe that science is better when it includes different perspectives and voices, and we are committed to elevating and nurturing scientists and other staff from traditionally excluded populations. We strive to make the Allen Institute a place where everyone feels like they belong and are empowered to do their best work in a supportive environment.

We strongly encourage people from all backgrounds to apply for our open positions.

Curious, innovative, and committed to finding answers and meaning in the complexities of bioscience. Regardless of how long Allenites stay with us, the knowledge gained here positions our people grow here or help advance human health elsewhere.

Explore our benefits guides for Washington Residents and Non-Washington Residents .

Explore career opportunities

Latest news

Our research

Science programs at allen institute.

• Share full article

Advertisement

Supported by

India Wins Cricket World Cup, Sealing Its Domination of the Sport

In India, cricket has become immensely profitable and a destination for the world’s best players. But a tournament victory had eluded it for many years.

By Mujib Mashal

Reporting from New Delhi

India won the men’s Cricket World Cup on Saturday, defeating South Africa to end a dry spell in tournament victories that had lasted over a decade, even as the nation was dominating the sport globally in other measures like talent, cash and influence.

The tournament was played across several Caribbean islands, with a few of the matches hosted in the United States, including at a pop-up stadium in New York. When the final, in Barbados, ended with India declared the champion, it was close to midnight back home, where joyful crowds poured into the streets across several cities.

“Maybe in a couple hours it will sink in, but it is a great feeling,” said Rohit Sharma, India’s captain, who took a tour of the stadium with his daughter propped on his shoulders to thank the crowd. “To cross the line — it feels great for everyone.”

It was a closely fought match, and a deeply emotional one for India, in part because many of its senior players, including Sharma, 37, were near the end of their careers. India last won the World Cup in T20, the shortest format of cricket, in 2007, when Sharma was just getting started. The top prize had also evaded Virat Kohli, 35, one of cricket’s most recognized icons. Rahul Dravid, India’s coach, had never won a World Cup during his long and illustrious career as a player.

All three men ended the night on a happy note, with Sharma and Kohli announcing their retirement from the fast-paced short form of the game. Dravid, who finished his stint as India’s coach, is normally a quiet, stoic presence. But after the win, he was screaming and celebrating.

We are having trouble retrieving the article content.

Please enable JavaScript in your browser settings.

Thank you for your patience while we verify access. If you are in Reader mode please exit and  log into  your Times account, or  subscribe  for all of The Times.

Thank you for your patience while we verify access.

Already a subscriber?  Log in .

Want all of The Times?  Subscribe .