Draft:Kim Stachenfeld
{{Short description|American neuroscientist and AI researcher}}
{{Draft topics|biography|stem}}
{{AfC topic|blp}}
{{AfC submission|||ts=20250512152844|u=Gehirnkarte|ns=118}}
{{AfC submission/draft}}
{{Infobox scientist
| honorific_prefix =
| name = Kim Stachenfeld
| honorific_suffix =
| native_name =
| native_name_lang =
| image =
| image_size =
| image_upright =
| alt =
| caption =
| birth_date =
| birth_place =
| death_date =
| death_place =
| death_cause =
| resting_place =
| resting_place_coordinates =
| other_names =
| residence =
| citizenship =
| nationality =
| fields = Neuroscience, AI
| workplaces = Google DeepMind, Columbia University
| patrons =
| education =
| alma_mater = Tufts University (B.S.), Princeton University (Ph.D.)
| thesis_title =
| thesis_url =
| thesis_year =
| doctoral_advisor = Matthew Botvinick
| academic_advisors =
| doctoral_students =
| notable_students =
| known_for =
| awards =
| author_abbrev_bot =
| author_abbrev_zoo =
| spouse =
| partner =
| children =
| signature =
| signature_alt =
| website =
| footnotes =
}}
Kimberly Lauren Stachenfeld is an American computational neuroscientist and artificial intelligence (AI) researcher. She serves as a Senior Research Scientist at Google DeepMind and holds an affiliate faculty position at the Center for Theoretical Neuroscience at Columbia University.{{Cite web |title=Kimberly L. Stachenfeld {{!}} Center for Theoretical Neuroscience |url=https://ctn.zuckermaninstitute.columbia.edu/people/kimberly-l-stachenfeld |access-date=2025-05-03 |website=ctn.zuckermaninstitute.columbia.edu}} She has made important contributions to the fields of neuroscience and machine learning, on how biological systems learn and represent information, and how these principles can inform AI development.{{Cite web |title=Kimberly Stachenfeld {{!}} Innovators Under 35 |url=https://www.innovatorsunder35.com/the-list/kimberly-stachenfeld/ |access-date=2025-05-03 |website=www.innovatorsunder35.com}}
Education
Stachenfeld earned dual bachelor's degrees in Mathematics and Chemical & Biological Engineering from Tufts University in 2013.{{Cite web |title=NeuroKim |url=https://neurokim.com/ |access-date=2025-05-03 |website=NeuroKim |language=en}} She then pursued a Ph.D. in Quantitative & Computational Neuroscience at Princeton University, completing it in 2018 under the supervision of Dr. Matthew Botvinick.{{Cite web|url=https://hai.stanford.edu/people/matthew-botvinick|title=Matthew Botvinick | Stanford HAI|website=hai.stanford.edu}} Her doctoral research centered on learning neural representations that support efficient reinforcement learning.{{Cite web |title=Learning Neural Representations that Support Efficient Reinforcement Learning - ProQuest |url=https://www.proquest.com/openview/90160aa5500d912a1fc6bfeb50acbf2c/1?cbl=18750&pq-origsite=gscholar |access-date=2025-05-03 |website=www.proquest.com |language=en}}
Research and Career
Stachenfeld's research explores the intersection of neuroscience and AI. In neuroscience, she investigates how animals construct and utilize internal models of their environment to support memory and prediction.{{Cite web |last1=Cepelewicz |first1=Jordana |last2=Magazine |first2=Quanta |date=2019-01-18 |title=A Hexagonal Theory of Memory |url=https://www.theatlantic.com/science/archive/2019/01/same-neurons-may-map-physical-space-and-memories/580357/ |access-date=2025-05-03 |website=The Atlantic |language=en}}{{Cite web |last=Cepelewicz |first=Jordana |date=2019-01-14 |title=The Brain Maps Out Ideas and Memories Like Spaces |url=https://www.quantamagazine.org/the-brain-maps-out-ideas-and-memories-like-spaces-20190114/ |access-date=2025-05-03 |website=Quanta Magazine |language=en}} In AI, she applies these insights to develop deep learning models that emulate cognitive functions.{{cite bioRxiv | biorxiv=10.1101/2025.02.05.636732 | title=Discovering Symbolic Cognitive Models from Human and Animal Behavior | date=2025 | last1=Castro | first1=Pablo Samuel | last2=Tomasev | first2=Nenad | last3=Anand | first3=Ankit | last4=Sharma | first4=Navodita | last5=Mohanta | first5=Rishika | last6=Dev | first6=Aparna | last7=Perlin | first7=Kuba | last8=Jain | first8=Siddhant | last9=Levin | first9=Kyle | last10=Éltető | first10=Noémi | last11=Dabney | first11=Will | last12=Novikov | first12=Alexander | last13=Turner | first13=Glenn C. | last14=Eckstein | first14=Maria K. | last15=Daw | first15=Nathaniel D. | last16=Miller | first16=Kevin J. | last17=Stachenfeld | first17=Kimberly L. }} She has contributed to projects involving reinforcement learning, graph neural networks,{{cite arXiv | eprint=2101.00079 | last1=Stachenfeld | first1=Kimberly | last2=Godwin | first2=Jonathan | last3=Battaglia | first3=Peter | title=Graph Networks with Spectral Message Passing | date=2020 | class=stat.ML }} and learned simulators for physical systems. Her work on predictive representations in the hippocampus has been influential in understanding how the brain anticipates future events.{{cite journal |doi=10.1038/nn.4650 |title=The hippocampus as a predictive map |date=2017 |last1=Stachenfeld |first1=Kimberly L. |last2=Botvinick |first2=Matthew M. |last3=Gershman |first3=Samuel J. |journal=Nature Neuroscience |volume=20 |issue=11 |pages=1643–1653 |biorxiv=10.1101/097170 |pmid=28967910 }}
Notable Works
- {{cite journal |doi=10.1038/nn.4650 |title=The hippocampus as a predictive map |date=2017 |last1=Stachenfeld |first1=Kimberly L. |last2=Botvinick |first2=Matthew M. |last3=Gershman |first3=Samuel J. |journal=Nature Neuroscience |volume=20 |issue=11 |pages=1643–1653 |biorxiv=10.1101/097170 |pmid=28967910 }}
References
{{Reflist}}
External links
- [https://ctn.zuckermaninstitute.columbia.edu/people/kimberly-l-stachenfeld Faculty profile at Columbia University]