Scholar
Nan Rosemary Ke
Google Scholar ID: dxwPYhQAAAAJ
Google, Deepmind, Mila
Citations & Impact
Citations
5,247
H-index
27
i10-index
38
Publications
20
Co-authors
13
list available
Contact
Publications
Resume (English only)
Academic Achievements
- - Published Papers:
- - Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities
- - Learning Beyond Pattern Matching? Assaying Mathematical Understanding in LLMs
- - Metacognitive Capabilities of LLMs: An Exploration in Mathematical Problem Solving
- - Scaling Instructable Agents Across Many Simulated Worlds
- - Learning to Induce Causal Structure
- - Sparse Attentive Backtracking: Temporal Credit Assignment Through Reminding
- - Focused Hierarchical RNNs for Conditional Sequence Processing
- - Twin Networks: Using the Future to Generate Sequences
- - Z-Forcing: Training Stochastic RNNs
- - Variational Walkback: Learning a Transition Operator as a Stochastic Recurrent Net
- - A Deep Reinforcement Learning Chatbot
- - Awards: Rising Star in Machine Learning, Rising Star in EECS, Facebook Fellowship recipient.
Research Experience
- - Staff Research Scientist at Google DeepMind, Key contributor: reasoning in Gemini series.
- - Contributor: reasoning in Gemini 2.5 and IMO efforts.
Education
- - Ph.D., Mila, Supervisors: Yoshua Bengio and Chris Pal
- - During her Ph.D., she also worked at Google DeepMind, Facebook/Meta AI Research, and Microsoft Research Montreal.
Background
- - Research Interests: Building reasoning systems for math, code, and decision-making.
- - Professional Field: Strengthening the reasoning capabilities of large models, causality, and modularity in deep learning.
- - Brief Introduction: Currently focusing on enhancing the reasoning capabilities of the Gemini family (including Gemini and Gemini 2.5). Previously, her research centered on causality and modularity in deep learning.
Miscellany
- - Personal Interests: Not provided
