Inference and Fine-tuning
Studying adaptation and inference-time behavior in modern machine learning systems.
MIT EECS · LIDS · IDSS
Kakei Yamamoto
山本かけい
PhD student in Computer Science at MIT, advised by Prof. Martin J. Wainwright. I work on inference, fine-tuning, generative models, machine learning theory, and optimization.
Research Assistant, Laboratory for Information & Decision Systems
Profile
Biography
I am a PhD student in Computer Science at MIT EECS and a research assistant at LIDS, advised by Martin J. Wainwright. I am also affiliated with IDSS.
My current research focuses on inference, fine-tuning, generative models, machine learning theory, and optimization, especially in settings involving distribution shift and model distillation.
Before MIT, I worked on deep learning theory at RIKEN AIP and studied Applied Mathematics and Computer Science at the University of Tokyo under Taiji Suzuki.
Interests
- Inference
- Distillation
- Fine-tuning
- Generative models
- Machine learning theory
- Optimization
Education
- PhD in Computer Science Massachusetts Institute of Technology · 2028 expected
- MS in Computer Science Massachusetts Institute of Technology · 2026
- BEng in Applied Mathematical and Computer Science The University of Tokyo · 2023
Awards
- Takenaka Oversea Scholarship Takenaka Scholarship Foundation, 2023-2028
- Kiyo Sakaguchi Scholarship Prudential Life Insurance Co., 2023-2027
Research
Questions I am currently drawn to.
Distribution Shift
Understanding when learning systems can recover from biased or misspecified supervision, especially in student-teacher estimation.
Mean-Field Methods
Analyzing Langevin dynamics, policy gradients, and distributional minimax optimization in regimes where feature learning matters.
Generative Modeling
Studying conditional diffusion models, classifier-free guidance, and the mathematical structure behind generative sampling.
Neural Computation
Exploring timing-based learning and sparse-firing regularization for spiking neural networks and efficient inference.
Publications
2026
TILT: Implicit Importance Weighting for Unsupervised Domain Adaptation
Submitted to NeurIPS, 2026
Introduces a one-step unsupervised domain adaptation framework that implicitly performs relative density-ratio reweighting without separately estimating density ratios.
2026
Residual-as-Teacher: Mitigating Bias Propagation in Student-Teacher Estimation
arXiv, 2026
Proposes residual-as-teacher estimation to reduce teacher-bias propagation and proves excess-risk and convergence guarantees.
2024
Mean Field Langevin Actor-Critic: Faster Convergence and Global Optimality beyond Lazy Learning
ICML, 2024
Develops mean-field Langevin actor-critic methods for feature-learning policy optimization.
2024
Symmetric Mean-field Langevin Dynamics for Distributional Minimax Problems
ICLR, 2024
Extends mean-field Langevin dynamics to minimax optimization over probability distributions.
2024
Can Timing-Based Backpropagation Overcome Single-Spike Restrictions in Spiking Neural Networks?
IJCNN, 2024
Studies timing-based backpropagation beyond single-spike restrictions in spiking neural networks.
2023
Sparse-firing regularization methods for spiking neural networks with time-to-first-spike coding
Scientific Reports, 2023
Introduces spike-timing-based regularization methods to reduce firing frequency in TTFS-coded spiking neural networks.
Portfolio
Selected Works
Released web game · 2026
BiteType! and BiteType! Classroom
Created and released a browser typing game set in a burger stand, plus a classroom mode for no-login school practice, teacher links, class results, and guided typing settings.
Community leadership · 2026
Japanese Association of MIT
Serving as JAM President. Led Hanami Festival as the representative organizer, planned Beneath the Great Wave from zero, and designed and built the JAM website.
Service
Leadership
Japanese Association of MIT (JAM)
President · 2025-
Representative organizer for Hanami Festival 2026; originated and planned Beneath the Great Wave; built the organization website and programming infrastructure.
BiteType!
Creator and developer · 2026
Designed, implemented, and released the typing game and its classroom product, including gameplay, audio, analytics, teacher controls, and Cloudflare deployment.
Teiyukai
Executive board member, General Affairs Director · 2022-2023
Served on the executive board of the University of Tokyo engineering student alumni/community organization.
UTokyo Philharmonic Orchestra
President · 2020-2021
Led the student orchestra while playing trumpet.