Tasha J. Kim

Other names: Jae Young, Bona

GitHub | LinkedIn | Stack Exchange | Email: tasha.kim [at] eng.ox.ac.uk

Educational Background

• Oxford University, Doctor of Philosophy (DPhil) - Engineering Science, '25-Current

• Stanford University, Master of Science (MS) - Computational and Mathematical Engineering, '25

• Stanford University, Graduate Program (certificate) - Artificial Intelligence, '22-'23

• Brown University, Bachelor of Science (ScB) - Mathematics, Computer Science, '21

Professional Background

• Stanford Institute for Human-centered Artificial Intelligence - Graduate Fellow, '24-'25

• Stanford Artificial Intelligence Laboratory - Graduate Researcher, '23-'25

• Google - Software Engineer, '22-'24

• National Institute of Standards and Technology - Research Associate, '21

Awards / Honors / Grants

• Oxford University Department of Engineering Science - Research grant, '25-'28

• Stanford University School of Engineering - Exceptional Graduate Student Award, '25

• Princeton University - Gordon Wu Fellowship, '25-'30 (declined)

• Stanford Institute for Human-centered Artificial Intelligence - Graduate Fellowship, '24-'25

• Google - Educational reimbursement (Stanford honors coop program), '23-'24

• Stanford Artificial Intelligence Alignment - AI Safety Policy Fellowship, '23

• Google - Educational reimbursement, '22-'23

• Brown University - The Albert A. Bennett Prize (Mathematics), '21

• National Institute of Standards and Technology - NIST PREP Fellowship, '21

• Major League Hacking Fellowship - Open Source Fellowship, '21

• Brown Office of The Vice President of Research - NIST Software Tool Development Fellowship, '20

• Brown University Harriet W. Sheridan Center - Research stipend, '20

• Brown University - Data Science Fellowship, '20

• Brown University - Karen T. Romer Undergraduate Teaching and Research Award (Mathematics), '18

• Brown University - Research stipend (Mathematics), '17

• The Duke of Edinburgh's Award - Gold Award, '16

Research Areas

1. Human-AI Cooperation. Effective human-AI cooperation requires systems that can understand goals, constraints and intent over time. By integrating persistent environment context, human signals, and machine learning, we can deliver intelligent systems that operate for critical real-world needs.

2. Interpretable Cognitive Reasoning. Reasoning in machines must be legible, inspectable, and revisable to be trustworthy. Instead of a hidden internal process, reasoning can be represented as a sequence of state transitions over memory. This allows agents to do more than just execute commands; they can justify decision-making, recover from errors, and evolve safely under human guidance.

3. Human-Aligned Autonomy. AI agents increasingly operate in high-stakes, long-horizon, shared environments. Hence, alignment cannot be a simple afterthought. Instead of relying on behavioral guarantees alone, alignment and governance primitives can be enforced at the infrastructure level. This enables trust, safety, and accountability as AI systems scale.

Talks / Workshops

• Oxford University Mathematical, Physical and Life Sciences Research Conference: AI & Ethics - Oxford, UK, '26 [p] [pr]

• AAAI: Advancing AI through Theory of Mind workshop - Singapore, '26 [p] [*]

• AAAI: Safe, Ethical, Certified, Uncertainty-aware, Robust, and Explainable AI workshop - Singapore, '26 [p] [*]

• International Joint Conference in Natural Language Processing & AACL - Mumbai, India, '25 [p] [pr] [*]

• NeurIPS: Embodied and Safe-Assured Robotics Systems workshop - New Mexico, Mexico, '25 [p] [pr] [*]

• NeurIPS: 2025 PNPL Competition - San Diego, CA, US, '25 [pr] [*]

• Stanford University ICME, 2025 Commencement - Stanford, CA, US, '25 [speech] 

• Stanford University School of Engineering, 100th Anniversary Centennial Celebration - Stanford, CA, US, '25 [demo] [pr]

• Stanford Institute for Human-centered AI - Stanford, CA, US, '25 [w]

• Stanford Vision and Learning Laboratory, Cognitive AI group - Stanford, CA, US, '25 [pr]

• Stanford Institute for Human-centered AI - Stanford, CA, US, '25 [pr]

• Stanford Intelligent Systems Laboratory - Stanford, CA, US, '25 [pr]

• Stanford University Institute for Computational and Mathematical Engineering - Stanford, CA, US, '25 [panel]

• Stanford AI Alignment, ARENA workshop - Stanford, CA, US, '25 [le] [w] 

• Stanford University ICME, 20th Anniversary Research Symposium - Stanford, CA, US, '24 [p]

• CoRL: Advancing Learning for Human-Centered Collaborative Robots workshop - Munich, Germany, '24 [pr] [spotlight] [*]

• Stanford AI Alignment, ARENA workshop - Stanford, CA, US, '24 [le] [w]

• Stanford Institute for Human-centered Artificial Intelligence, AI4All - Stanford, CA, US, '24 [le]

• Stanford Intelligent Systems Laboratory - Yosemite, CA, US, '24 [pr]

• Google, Knowledge and Information - Mountain View, CA, US, '23 [pr]

• Meta, Privacy Infrastructure team - Menlo Park, CA, US, '22 [pr]

• Google, Core Data Storage team - New York, NY, US, '22 [pr]

• Major League Hacking Fellowship - US, '21 [pr][*]

• NASA Lucy Student Pipeline Accelerator and Competency Enabler - US, '21 [pr] [*]

• Brown University Sheridan Center - Providence, RI, US, '20 [pr]

• Joint Mathematics Meetings - MD, US, '19 [p] [-]

• The Institute for Computational and Experimental Research in Mathematics - Providence, RI, US, '18 [pr]

• Brown University Science Center - Providence, RI, US, '18 [p]

• Brown University Department of Mathematics - Providence, RI, US, '17 [pr]

• Jewish Community Day School of Rhode Island - Providence, RI, US, '17 [w]

[w]: workshop, [pr]: presentation, [p]: poster, [le]: lecture, [*]: remote, [-]: absent

Academic Service

• Acknowledged Contributor, Algorithms for Validation (MIT Press, forthcoming; preview '25): feedback on early manuscript draft and contribution to accessibility alt text.

• Acknowledged Contributor, Origamis Associated with Minimally-Intersecting Filling Pairs (Pacific Journal of Mathematics, '22): research contributions during 2018 Summer @ ICERM program supporting the paper's experimental results.

• Reviewer:

  • Journals: IEEE Robotics and Automation Letters (RA-L) '26.

  • Conferences: HRI '26 alt HRI; CHI '26.

  • Workshops: CoRL '24; NeurIPS '24, '25; RSS '25; ICLR '25; AAAI '26.

  • Other: HRI '26 Late-Breaking Reports; IEEE Robotics and Automation Magazine '25.

• Dean's Graduate Student Advisory Council - Stanford University School of Engineering, '24-'25

• Executive Leader '25, Project Lead '24 - Stanford Artificial Intelligence Alignment, '24

• Student Representative - Stanford University Institute for Computational and Mathematical Engineering Student Action Group, '24-'25

• Lead Mentor '25, Mentor '24 - AI, Robotics, Stanford Institute for Human-centered Artificial Intelligence (HAI), AI4All.

• Founder '24, Co-Lead '24-'25 - Cognitive Machine Intelligence HAI Affinity Group.

• Student Leader '20-'21, Officer '17-'21 - Brown University Mathematics Departmental Undergraduate Group.

• Organizing Committee - The Symposium for Undergraduates in the Mathematical Sciences, '17, '18.

Professional Affiliations

• Institute of Electrical and Electronics Engineers (IEEE), '23-Current

• Society of Industrial and Applied Mathematics (SIAM), '23-Current