1 citations · 2 across the 10 of their papers we have counts for
8 papers · 1 filter
Periodic Weak Spots: Phase Sensitivity from Chunked KV-Cache Compression
Xingyu Zhu, Pu, Yi +6
Chunked KV-cache compression reduces the memory and attention costs of long-context inference by compressing windows of consecutive tokens into fewer cache entries at a fixed strid…
Aligning One-Step Generative Models with Reward-Weighted Transport Distillation
Austin Wang, Ziheng Cheng, Lexing Ying
One-step generators enable high-quality visual generation with a single network evaluation, but their post-training is difficult: general implicit generators provide neither tracta…
DriftLite: Lightweight Drift Control for Inference-Time Scaling of Diffusion Models
Yinuo Ren, Wenhao Gao, Lexing Ying +2
We study inference-time scaling for diffusion models, where the goal is to adapt a pre-trained model to new target distributions without retraining. Existing guidance-based methods…
Solving Inverse Problems via Diffusion-Based Priors: An Approximation-Free Ensemble Sampling Approach
Haoxuan Chen, Yinuo Ren, Martin Renqiang Min +2
Diffusion models (DMs) have proven to be effective in modeling high-dimensional distributions, leading to their widespread adoption for representing complex priors in Bayesian inve…
A Unified Approach to Analysis and Design of Denoising Markov Models
Yinuo Ren, Grant M. Rotskoff, Lexing Ying
Probabilistic generative models based on measure transport, such as diffusion and flow-based models, are often formulated in the language of Markovian stochastic dynamics, where th…
COS-DPO: Conditioned One-Shot Multi-Objective Fine-Tuning Framework
Yinuo Ren, Tesi Xiao, Michael Shavlovsky +2
In LLM alignment and many other ML applications, one often faces the Multi-Objective Fine-Tuning (MOFT) problem, i.e., fine-tuning an existing model with datasets labeled w.r.t. di…
Understanding the Generalization Benefits of Late Learning Rate Decay
Yinuo Ren, Chao Ma, Lexing Ying
Why do neural networks trained with large learning rates for a longer time often lead to better generalization? In this paper, we delve into this question by examining the relation…
Multi-Objective Optimization via Wasserstein-Fisher-Rao Gradient Flow
Yinuo Ren, Tesi Xiao, Tanmay Gangwani +4
Multi-objective optimization (MOO) aims to optimize multiple, possibly conflicting objectives with widespread applications. We introduce a novel interacting particle method for MOO…