Robust Information-Gain Control
Fixing belief trap failures in LLM agents with distributionally robust information gain
My works
These are projects conducted in my classes at MIT and beyond
What Matters in RL for Diffusion Models?
The dominant role of noise in RL post-training for diffusion models
Improving Activation Steering
Gated Cropped Attention-Delta Steering fixes KV-cache contamination in multi-turn dialogue
KINA Benchmark
A high-density benchmark spanning 261 disciplines with game-theoretic annotation
CogGym
A scalable framework for comparing AI models against humans using cognitive science experiments
40Hz Toolbox
VR-based 40Hz light therapy for adult amblyopia and Alzheimer's disease
Caveman Papers
Animation agents that turn research papers into short-form reels
Persona Collapse
Why many “different” AI personas end up sounding strangely the same
Say Something Else
Privacy-preserving LLM communication through information sufficiency
Rewarded Region Replay (R3)
Reinforcement Learning in Sparse Reward Environments
The Temporal Cave
Testing the Platonic Representation Hypothesis for video
Texture Synthesis and Transfer
Final Project for 6.4400 Computer Graphics
Hanoiiwa
Robotics Manipulation
Is She Smiling?
Testing bias in facial expression recognition
Cryptography and Security
Analyzing Memory Usage of Adversarially Resistant Bloom Filters
Localziation, Path Planning and Lane Racing
Robotics Science and Systems
Trashinator
A robot that picks up trash
Team Mulan
The first all-girls robotics team in China
Mini PID bot
NEET Autonomous Machines Competition