Kianté Brantley

(He/Him/His)
Talks

Assistant Professor
Kempner Institute and SEAS,
Harvard University
150 Western Av., Allston MA 02134

[kdbrantley@g.harvard.edu] [CV]

Invited Talks:

2026

Regression as Policy Optimization: Advantages In, Policies Out
IMSI Frontiers in Online Reinforcement Learning, 2026
Mid-Atlantic Student Colloquium on AI, Language and Learning, 2026
Power & AI Initiative Research Summit, 2026
[Slides]
Train, Reason, and Act Under Value Function Guidance
University of Maryland, College Park, 2026
[Slides]

2025

Coactive Learning for Large Language Modes using Implicit User Feedback
Informs Session: Preference Learning in Large Language Models, 2025
[Slides]
The Power of Resets!
CCC Computing Futures Symposium, 2025
Google, 2025
Berkeley Simons Institute: The Future of Language Models and Transformers, 2025
[Slides]
Efficient Policy Optimization Techniques for LLMs
NSF RL Workshop Panel, 2025
[Slides]

2024

Efficient Policy Optimization Techniques for LLMs
MIT, 2024
[Slides]
Learning from Interaction
George Washington University, 2024
Emory University, 2024
Stony Brook University, 2024
Columbia University, 2024
Rice University, 2024
Harvard University, 2024
Rutgers University, 2024
Princeton University, 2024
Microsoft Research - NYC, 2024
University of Michigan, 2024
Northeastern, 2024
University of Illinous Chicag, 2024
[Slides]

2023

Reinforcement Learning from Guided Feedback: Addressing the Shortcoming of RL in NLP
Rochester Institute of Technology, 2023
NC State University, 2023
[Slides]
Ranking with Long Term Constraints
Wayfair, 2023
[Slides]
Expert-in-the-Loop for Sequential Decisions and Predictions
Cornell NLP Seminar, 2023
Google Brain Montreal, 2023
[Slides]

2021

Expert-in-the-Loop for Sequential Decisions and Predictions
University of Maryland UMIACS, 2021
[Slides]
Successor feature sets: Generalizing successor representations across policies
Microsoft Research Summit, 2021
[Slides]

2020

Learning through interaction with experts
Adobe Research, 2020
[Slides]
Active Imitation Learning with Noisy Guidance
The New York Academy of Sciences, 2020
[Slides]
Non-monotonic sequential text generation
Black in AI Workshop at ACL, 2020
[Slides]