Meta tags:
description=I am a Research Scientist at Google DeepMind. br br My research has focused on i closed-loop methods /i for i efficient adaptation /i , using a model s own performance as signal for improvement. This has spanned targeted training-data generation and gradient-free test-time adaptation, and, more recently, self-improving models. br br My PhD was at b EPFL /b supervised by i Amir Zamir /i , on making models more reliable under changing environments. I was also a postdoc at the b Singapore-MIT /b research centre working on neurosymbolic methods for adaptation. In my past life, I was a quant in New York and London and worked on creating systematic investment strategies (or, a glorified coin flipper). ;

Headings (most frequently used words):

data, models, of, learning, test, time, and, workshop, co, organizer, robustness, via, adaptation, adapt, neural, networks, task, the, cross, teaching, epfl, technology, teresa, yeo, controlled, training, generation, with, diffusion, an, analysis, model, across, concurrent, distribution, shifts, viper, visual, personalization, generative, individual, preference, 4m, massively, multimodal, masked, modelling, rapid, network, to, using, feedback, discovery, finding, tasks, that, generalize, on, 3d, common, corruptions, augmentation, domain, ensembles, robust, through, consistency, iterative, classroom, university, cambridge, postdoctoral, researcher, singapore, mit, alliance, for, research, assistant, scientist, shift, quantitative, researher, ubs, reviewer, 3rd, updates, catch, operate, monitoring, ml, under, drift,

Text of the page (most frequently used words):
yeo (13), and (13), zamir (10), paper (10), code (9), for (8), project (8), page (8), #models (7), research (6), 2026 (6), #adaptation (6), learning (6), test (5), time (5), kar (5), 2025 (4), iclr (4), 2023 (4), data (4), oral (4), demo (4), teresa (3), workshop (3), organizer (3), adapt (3), the (3), neurips (3), cvpr (3), iccv (3), scientist (3), machine (3), 2019 (3), epfl (3), methods (3), singapore (3), 2024 (3), atanov (3), was (3), with (2), catch (2), operate (2), under (2), updates (2), eccv (2), systematic (2), strategies (2), 2015 (2), technology (2), 2016 (2), 2017 (2), visual (2), intelligence (2), fall (2), 2018 (2), 2020 (2), 2021 (2), teaching (2), neurosymbolic (2), efficient (2), mit (2), thesis (2), making (2), computer (2), robust (2), amir (2), dillenbourg (2), sax (2), cross (2), task (2), robustness (2), via (2), 2022 (2), neural (2), networks (2), using (2), bachmann (2), choi (2), model (2), tmlr (2), training (2), generation (2), google (2), deepmind (2), has (2), more (2), built, website, template, likes, dogs, icml, monitoring, drift, 3rd, reviewer, present, researched, equity, portfolios, quantitative, researher, ubs, 2013, designed, impelmented, automated, fraud, detection, shift, cs503, machines, minds, ee559, deep, cs433, spring, assistant, postdoctoral, researcher, alliance, bayesian, optimization, natural, language, processing, phil, university, cambridge, vision, adaptive, advisor, pierre, science, kamalaruban, singla, merchant, asselborn, faucon, cevher, iterative, classroom, aaai, cheerla, suri, cao, malik, guibas, through, consistency, arxiv, domain, ensembles, common, corruptions, augmentation, filatov, sohmshetty, discovery, finding, tasks, that, generalize, sodagar, rapid, network, feedback, mizrahi, gao, dehghan, massively, multimodal, masked, modelling, spotlight, salehi, shafiei, viper, personalization, generative, individual, preference, jeon, analysis, across, concurrent, distribution, shifts, benoit, alekseev, ray, esmaeil, akhoondi, controlled, diffusion, our, workshops, have, been, accepted, see, you, rio, dec, started, feb, focused, own, performance, signal, improvement, this, spanned, targeted, gradient, free, recently, self, improving, phd, supervised, reliable, changing, environments, also, postdoc, centre, working, past, life, quant, new, york, london, worked, creating, investment, glorified, coin, flipper, closed, loop, services, experience, education, publication, researching, things,

Text of the page (random words):
teresa yeo teresa yeo researching things publication education experience services i am a research scientist at google deepmind my research has focused on closed loop methods for efficient adaptation using a model s own performance as signal for improvement this has spanned targeted training data generation and gradient free test time adaptation and more recently self improving models my phd was at epfl supervised by amir zamir on making models more reliable under changing environments i was also a postdoc at the singapore mit research centre working on neurosymbolic methods for adaptation in my past life i was a quant in new york and london and worked on creating systematic investment strategies or a glorified coin flipper feb 2026 i started as a research scientist at google deepmind in singapore dec 2025 our workshops on test time updates and catch adapt and operate have been accepted at iclr 2026 see you in rio tmlr 2025 controlled training data generation with diffusion models t yeo a atanov h benoit a alekseev r ray p esmaeil akhoondi a zamir project page paper code tmlr 2025 an analysis of model robustness across concurrent distribution shifts m jeon s choi h choi t yeo paper code eccv 2024 viper visual personalization of generative models via individual preference learning s salehi m shafiei r bachmann t yeo a zamir project page paper code demo neurips 2023 spotlight 4m massively multimodal masked modelling d mizrahi r bachmann o f kar t yeo m gao a dehghan a zamir project page paper code demo iccv 2023 rapid network adaptation learning to adapt neural networks using test time feedback t yeo o f kar z sodagar a zamir project page paper code neurips 2022 task discovery finding the tasks that neural networks generalize on a atanov a filatov t yeo a sohmshetty a zamir project page paper code cvpr 2022 oral 3d common corruptions and data augmentation o f kar t yeo a atanov a zamir project page paper code demo iccv 2021 oral robustness via cross domain ensembles t yeo o f kar a sax a zamir project page paper code arxiv cvpr 2020 oral robust learning through cross task consistency a zamir a sax t yeo o f kar n cheerla r suri z cao j malik l guibas project page paper code demo aaai 2019 oral iterative classroom teaching t yeo p kamalaruban a singla a merchant t asselborn l faucon p dillenbourg v cevher paper 2017 2024 epfl ph d in computer science advisor amir zamir pierre dillenbourg thesis making computer vision models robust and adaptive 2015 2016 university of cambridge m phil in machine learning and machine intelligence thesis bayesian optimization for natural language processing 2024 2026 postdoctoral researcher singapore mit alliance for research and technology neurosymbolic methods for efficient adaptation 2018 2023 teaching assistant epfl fall 2021 cs503 visual intelligence machines and minds spring 2018 2019 2020 ee559 deep learning fall 2019 cs433 machine learning 2016 2017 data scientist shift technology designed and impelmented models for automated fraud detection 2013 2015 quantitative researher ubs researched on systematic strategies for equity portfolios 2023 present reviewer neurips iclr cvpr iccv eccv 2026 the 3rd test time updates workshop co organizer iclr 2026 catch adapt and operate monitoring ml models under drift workshop co organizer iclr 2025 test time adaptation workshop co organizer icml 2026 teresa yeo likes dogs built with research website template