Lena Strobl, Ph.D.
Postdoctoral Associate in Linguistics at Yale University
I study what language models can compute, and lately, their social life.
I love teaching, and I love organizing. Seminars, workshops, whole interdisciplinary programs, and the communities that grow out of them.
About me
Positions
- Postdoctoral Associate, Yale LinguisticsCLAY Lab, with Robert Frank, since 2026
- Assistant Researcher, Yale2024
- Researcher, MIT2023–2024
- Research Intern, Allen Institute for AIAristo group, 2023
Education
- PhD in Computer Science, Umeå UniversityThesis: Transformers as Recognizers and Transducers, 2025
- MSc with Distinction, University of SheffieldComputer Science with Speech and Language Processing, 2021
- BSc, Freie Universität BerlinComputer Science and Mathematics, 2019
Awards
- Knut and Alice Wallenberg Foundation postdoctoral fellowshipfunds the Yale postdoc
- Amazon Prize for Best Overall PerformanceTop of the 2021 MSc class, University of Sheffield
Papers
- 2026Simulating Hard Attention Using Soft AttentionAndy Yang, Lena Strobl, David Chiang, Dana Angluin. Transactions of the ACL 14
- 2026Concise One-Layer Transformers Can Perform Dictionary LookupAnton Hampe, Lena Strobl, Robert Frank, Dana Angluin. ICGI
- 2025Transformers as TransducersLena Strobl, Dana Angluin, David Chiang, Jonathan Rawski, Ashish Sabharwal. Transactions of the ACL 13
- 2024What Formal Languages Can Transformers Express? A SurveyLena Strobl, William Merrill, Gail Weiss, David Chiang, Dana Angluin. Transactions of the ACL 12
- 2024Computational Expressivity of Neural Language ModelsButoi, Chan, Cotterell, Merrill, Nowak, Pasti, Strobl, Svete. ACL tutorial
- 2023
Teaching, talks & service
- Spring 2027Formal Languages and Neural Networks, course at Yale University. Co-developer, with Robert Frank
- 2024Expressivity of Transformers: Logic, Circuits, and Formal Languages, course at ESSLLI, Leuven. Lecturer
- 2024Computational Expressivity of Neural Language Models, tutorial at ACL, Bangkok. Co-author and presenter
- 2022–2023Computational Complexity, Umeå University. Assisting lecturer and head teaching assistant
- 2022Efficient Algorithms, Umeå University. Lecturer
- 2020Logic and Discrete Mathematics, block course, FU Berlin. Teaching assistant
- 2020Foundations of Theoretical Computer Science, FU Berlin. Lead teaching assistant and mentor
- 2020Codes, Ciphers and Encrypted Messages, summer course for high-school students, FU Berlin. Mentor
- 2019Linear Algebra, FU Berlin. Teaching assistant
- 2018–2019Algorithms, Data Structures and Data Abstraction, FU Berlin. Lead teaching assistant
- 2018Logic and Discrete Mathematics, block course, FU Berlin. Teaching assistant
- 2016–2019Girls' Day, introducing high-school girls to computer science, FU Berlin. Mentor
- 2025Transformers as Transducers, Mathematics of Language (MOL), Stony Brook. Invited talk
- 2025Transformers as Transducers, ACL 2025, Vienna. Paper presentation
- 2024Transformers as Recognizers and Transducers, Yale University. Guest lecture
- 2024Transformers and Circuits, Boston College, Howard Straubing's group. Talk
- 2024Transformers as Recognizers and Transducers, Google DeepMind, Heiga Zen's group, Tokyo. Talk
- 2024The Expressive Power of Attention-Based Models: Bridging Machine Learning and Logic, Logic and AI Workshop, Amsterdam. Invited talk
- 2024What Formal Languages Can Transformers Express? A Survey, NAACL, Mexico City. Poster
- 2023Theoretical Insights into Transformer Clockworks, Saarland University, Alexander Koller's and Michael Hahn's groups, Saarbrücken. Two talks
- 2023Theoretical Expressivity of Transformers and Their Generative Potential, MIT, Roger Levy's lab and Bob Berwick's group at CSAIL. Talk and discussion
- 2023Formal Language Theoretic Expressivity of Transformers, CSAIL, MIT. Poster
- 2018Shellability, Einstein Institute International Undergraduate Workshop, Hebrew University of Jerusalem. Invited talk
- 2026FLaNN Workshop, Yale University. Proposer and co-organizer
- 2025Dagstuhl Seminar 25282, Theory of Neural Language Models, Wadern. Co-organizer, one of four
- 2025What Can('t) Transformers Do?, NeurIPS workshop, San Diego. Co-organizer, one of four
- Since 2024International Conference on Grammatical Inference (ICGI). Steering committee member
- 2024Expressivity of Transformers: Logic, Circuits, and Formal Languages, course at ESSLLI, Leuven. Proposer, co-organizer, and lecturer
- 2024Computational Expressivity of Neural Language Models, tutorial at ACL, Bangkok. Co-author and presenter
- 2024WASP study trip to Japan, visiting Google DeepMind, RIKEN, NII, and Kyoto and Nagoya Universities. Organizer
- 2023DLT and WORDS conferences, Umeå. Co-organizer
- 2023Foundations of Language Processing Workshop, Granö. Co-organizer
- Since 2022FLaNN seminar, a weekly online seminar on formal languages and neural networks. Host and organizer
- 2022–2023PhD student representation, Umeå University. Co-chair, PhD Student Council. Vice-chair, WASP Graduate School council
- 2020–2021Student representation, University of Sheffield. Student representative on the Faculty of Engineering council and the Computer Science Staff Student Liaison Committee
- 2019–2020Student representation, FU Berlin. Chair, Teaching and Learning Committee. Student representative on the Faculty Council and the Teaching Assistant Committee. Member of hiring committees for professorships and of committees on study programs and teaching
Writing
Ideas from my research and beyond, explained with doodles, illustrations, and animations.
Explore the writing