Blog
Blog
Research notes, project writeups, and blogpost on interesting findings.
Struggling Toward Generative Model Evaluation
A blog post on why generative model evaluation is hard, how representation-based metrics became the default language, and why benchmarks are useful but never final.
Taming Outlier Tokens in Diffusion Transformers
Outlier patch tokens hurt both ViT encoders and diffusion transformers in RAE-DiT pipelines. Our Dual-Stage Registers (DSR) patch both sides and improve ImageNet-256 FID from 5.89 → 4.58 at 80 epochs.
EMA: A Quiet Hyperparameter That Moves Diffusion Leaderboards
We benchmark EMA decay across pixel-, latent-, and representation-space diffusion models and show that the popular 0.9999 default silently trades recall for precision.
Uni-Instruct: One-step Diffusion through Unified Divergence Instruction
A single f-divergence framework that subsumes 10+ one-step diffusion distillation methods — and a new SoTA one-step FID of 1.02 on ImageNet 64×64. NeurIPS 2025.